Microstrip Antennas PDF

International Journal of Antennas and Propagation
Microstrip Antennas
Microstrip Antennas
Microstrip Antennas
Copyright © 2012 Hindawi Publishing Corporation. All rights reserved.
This is a focus issue published in “International Journal of Antennas and Propagation.” All articles are open access articles distributed
under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, pro-
vided the original work is properly cited.
Editorial Board
M. Ali, USA Se-Yun Kim, Republic of Korea Sadasiva M. Rao, USA
Charles Bunting, USA Ahmed A. Kishk, Canada Sembiam R. Rengarajan, USA
Felipe Cátedra, Spain Tribikram Kundu, USA Ahmad Safaai-Jazi, USA
Dau-Chyrh Chang, Taiwan Byungje Lee, Republic of Korea Safieddin Safavi-Naeini, Canada
Deb Chatterjee, USA Ju-Hong Lee, Taiwan Magdalena Salazar-Palma, Spain
Z. N. Chen, Singapore L. Li, Singapore Stefano Selleri, Italy
Michael Yan Wah Chia, Singapore Yilong Lu, Singapore Krishnasamy T. Selvan, India
Christos Christodoulou, USA Atsushi Mase, Japan Zhongxiang Q. Shen, Singapore
Shyh-Jong Chung, Taiwan Andrea Massa, Italy John J. Shynk, USA
Lorenzo Crocco, Italy Giuseppe Mazzarella, Italy M. Singh Jit Singh, Malaysia
Tayeb A. Denidni, Canada Derek McNamara, Canada Seong-Youp Suh, USA
Antonije R. Djordjevic, Serbia C. F. Mecklenbräuker, Austria Parveen Wahid, USA
Karu P. Esselle, Australia Michele Midrio, Italy Yuanxun Ethan Wang, USA
Francisco Falcone, Spain Mark Mirotznik, USA Daniel S. Weile, USA
Miguel Ferrando, Spain Ananda S. Mohan, Australia Quan Xue, Hong Kong
Vincenzo Galdi, Italy P. Mohanan, India Tat Soon Yeo, Singapore
Wei Hong, China Pavel Nikitin, USA Young Joong Yoon, Korea
Hon Tat Hui, Singapore A. D. Panagopoulos, Greece Wenhua Yu, USA
Tamer S. Ibrahim, USA Matteo Pastorino, Italy Jong Won Yu, Republic of Korea
Nemai Karmakar, Australia Massimiliano Pieraccini, Italy Anping Zhao, China
Contents
Modal Resonant Frequencies and Radiation Quality Factors of Microstrip Antennas, Jan Eichler,
Pavel Hazdra, Miloslav Capek, and Milos Mazanek
Volume 2012, Article ID 490327, 9 pages
Tunable Compact UHF RFID Metal Tag Based on CPW Open Stub Feed PIFA Antenna, Lingfei Mo and
Chunfang Qin
Some Recent Developments of Microstrip Antenna, Yong Liu, Li-Ming Si, Meng Wei, Pixian Yan,
Pengfei Yang, Hongda Lu, Chao Zheng, Yong Yuan, Jinchao Mou, Xin Lv, and Housjun Sun
New Configurations of Low-Cost Dual-Polarized Printed Antennas for UWB Arrays, Guido Valerio,
Simona Mazzocchi, Alessandro Galli, Matteo Ciattaglia, and Marco Zucca
Design and Analysis of Wideband Nonuniform Branch Line Coupler and Its Application in a Wideband
Butler Matrix, Yuli K. Ningsih, M. Asvial, and E. T. Rahardjo
Isolation Improvement of a Microstrip Patch Array Antenna for WCDMA Indoor Repeater Applications,
Hongmin Lee and Jinwon Park
Series-Fed Microstrip Array Antenna with Circular Polarization, Tuan-Yung Han

Vertical Meandering Approach for Antenna Size Reduction, Li Deng, Shu-Fang Li, Ka-Leung Lau,
and Quan Xue
Microstrip Patch Antenna Bandwidth Enhancement Using AMC/EBG Structures, R. C. Hadarig,

M. E. de Cos, and F. Las-Heras
High-Performance Computational Electromagnetic Methods Applied to the Design of Patch Antenna

with EBG Structure, R. C. Hadarig, M. E. de Cos, and F. Las-Heras
A Wideband High-Gain Dual-Polarized Slot Array Patch Antenna for WiMAX Applications in 5.8 GHz,
Amir Reza Dastkhosh and Hamid Reza Dalili Oskouei
Hindawi Publishing Corporation
doi:10.1155/2012/490327
Research Article
Modal Resonant Frequencies and Radiation Quality Factors of
Microstrip Antennas
Jan Eichler, Pavel Hazdra, Miloslav Capek, and Milos Mazanek

Department of Electromagnetic Field, Faculty of Electrical Engineering, Czech Technical University in Prague, Technická 2,
166 27, Prague, Czech Republic
Correspondence should be addressed to Pavel Hazdra, hazdrap@fel.cvut.cz
Received 9 August 2011; Revised 10 January 2012; Accepted 13 January 2012
Academic Editor: Charles Bunting
Copyright © 2012 Jan Eichler et al. This is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
The chosen rectangular and fractal microstrip patch antennas above an infinite ground plane are analyzed by the theory of
characteristic modes. The resonant frequencies and radiation Q are evaluated. A novel method by Vandenbosch for rigorous
evaluation of the radiation Q is employed for modal currents on a Rao-Wilton-Glisson (RWG) mesh. It is found that the
resonant frequency of a rectangular patch antenna with a dominant mode presents quite complicated behaviour including having
a minimum at a specific height. Similarly, as predicted from the simple wire model, the radiation Q exhibits a minimum too. It is
observed that the presence of out-of-phase currents flowing along the patch antenna leads to a significant increase of the Q factor.
1. Introduction Equation (1) is usually treated within the method of

moments (MoMs) [8] framework and, due to the structure
Evaluation of the basic properties of microstrip patch discretization, the L operator is known as the “complex
antennas (MPA) has been numerously discussed in literature, impedance matrix” [Z] = [R] + j[X].
see, for example, [1–3]. The two main MPA attributes are Then the associated Euler’s equation to be solved is
resonant frequency (or frequencies but we will deal mostly
with the dominant mode) and the radiation Q factor. So far, XJn = λn RJn . (2)
only approximate results and semianalytic equations have
been published. To our knowledge, this is the first time
Equation (2) is a standard weighted eigenvalue equation
that these important characteristics have been studied in a
leading to a set of real characteristic eigencurrents Jn and
rigorous way. The antennas are treated by using a modal
associated eigenvalues λn . Properties of eigenvalues are
approach (hence we do not a priori consider any feeding
described in [9], at this moment it is important to note
to be connected), namely, by the theory of characteristic
that λn reflects the amount of net reactive power (thus λn =
modes (TCM), [4, 5]. Evaluation of the radiation Q is
0 means resonance). Instead of eigenvalues, the so-called
performed both by the TCM from the eigenvalues slope and
characteristic angles αn are introduced to show more visible
by novel rigorous equations derived by Vandenbosch [6] and
behavior with frequency [9]. Characteristic currents form a
Vandenbosch and Volski [7].
complete orthogonal set, and hence the total current on a
conducting body may be expressed as a linear combination
2. The Theory of Characteristic Modes of these mode currents [10].
For completeness, let us formulate the basics of the charac-
teristic modes for perfectly conducting bodies of area S. The 2.1. Implementation of the Characteristic Modes Theory.
scattered field Es is related to the electric surface currents J by Implementation of the modal decomposition process has
the electric field integral equation (EFIE) [8] been done in the MATLAB [11] environment using Makarov
EFIE codes [12] with the RWG basis functions [13]. This
L(J) − Ei tan
= 0. (1) usage is restricted to arbitrary 3D PEC structures with air
2 International Journal of Antennas and Propagation
dielectrics. Our developed TCM tool [14] has the following

T1 (x,y,z)
main advantages:
(i) Comsol Multiphysics [15]/MATLAB’s PDE TooLbox J1
mesh import, r21 = |r2 − r1 |
Ω
(ii) Optional Green’s function for infinite ground plane
simulations, J2
(iii) Single solver/multicore solver/distributed solver r2 r1
y
(within a computer network with installed
MATLAB). (0,0) T2 (x,y,z)
x
3. The Radiation Q Factor Figure 1: Distance between nonoverlapping current elements [23].
In [6] a novel theory able to rigorously calculate radiated

power and stored energies directly from currents flowing In [18], (10) is supposed to be an approximation of the
along the antenna has been presented. The radiation Q factor radiation Q, but in resonance it is actually exact.
is then readily evaluated by the definition [16]: Since characteristic modes are normalized to radiate unit
power Pr = 1 [4], (3) reduces to
m , W
max W e
Q = 2ω . (3)
Pr e , W
Q = 2ω max W m . (11)
The equations for radiated power Pr and stored electric and For parallel or series RLC circuit (hence, for one mode), the
e , W
magnetic energies W m are
“impedance QZ ” equals the exact “current Q” [6]:

1

Pr = k2 J(r1 )J(r2 ) − ∇ · J(r1 )∇ · J(r2 ) ω0
∂Z = ω0 ∂R + ∂X .
8πωε0 Q = QZ = (12)
2 ∂ω 2 ∂ω ∂ω
Ω1 Ω2
sin(kr21 )
× dΩ1dΩ2,
r21 Inserting
(4)
1
m − W
e

Z = R + jX = Pr + j2ω W (13)
|I |
2
e = 1
W (Ie − IR ), (5)
16πω2 ε0 valid for lossless antennas [19] and using the fact that Pr = 1,
(12) results in
m = 1
W (Im − IR ), (6)
16πω2 ε0 ω0
∂X = ω0 ∂ 2ω Wm − W
e
Q = QX =
where 2 ∂ω 2 ∂ω
(14)
ω0 ∂λ
k
= = Qeig ,
IR = k2 J(r1 )J(r2 ) − ∇ · J(r1 )∇ · J(r2 ) 2 ∂ω
2 Ω1 Ω2 (7)
× sin(kr21 )dΩ1dΩ2, providing that

m − W
λ = 2ω W e . (15)
cos(kr21 )
Ie = ∇ · J(r1 )∇ · J(r2 ) dΩ1dΩ2, (8)
Ω1 Ω2 r21 It is therefore concluded that the modal Qeig equals the QX by
definition, and it can be proven (using the reactance theorem
cos(kr21 )
Im = k 2 J(r1 )J(r2 ) dΩ1dΩ2, (9) [20, 21]) that in resonance it also equals the radiation Q
Ω1 Ω2 r21
defined from energies by (11).
where k is a free-space wavenumber, J is the surface current
density, and r21 is the distance between interacting current 3.2. Software Implementation. The above equations were
elements. The tilde denotes that the radiation contribution IR implemented in MATLAB for the RWG triangular mesh
has been subtracted from the stored energies at every point where two different interaction situations occur:
in space [17]. It is assumed that the currents are flowing in a
vacuum.
(a) Distant Elements. When the triangular elements are not
overlapping, current density on triangles may be simply
3.1. The Modal Radiation Q Factor. The modal radiation Q approximated as point sources located at the centre of
factor may be evaluated from the slope of modal eigenvalues triangles [22], see Figure 1. No actual integration is then
[18]: needed. This centroid approach is very fast with satisfactory
ω0 dλ accuracy as will be shown later (however it may fail for
Qeig = . (10) patches located very close to the ground plane).
2 dω
International Journal of Antennas and Propagation 3
P3 (x3 , y3 ) z
β
h13 (0,1)
A T1 (x,y,z) T1 (α ,β,γ) 0.01
P1 (x1 , y1 ) P0
P0
h23 α
0.005
r2
y h12 (0,0) (1,0) 0
P2 (x2 , y2 ) 2H
r1
(0,0) x
−0.005
y
(a) (b)
Figure 2: Self-term evaluation. (a) Original problem, (b) simplex −0.01 x
coordinates transformation [23].
0.015
0.01
(b) Overlapping (Self) Elements. As known from the method 0.005

of moments, the so-called “self ” contributions are of great 0
importance when dealing with calculations on discrete −0.005
elements (meshes). 0.02
−0.01 0.01
Here, the self-interaction occurs when two triangles are 0
−0.015 −0.01
−0.02
overlapping each other. Due to the behavior of integral −0.03
kernels, only rapidly varying term cos(kr21 )/r21 has to be
carefully treated. Since k0 R21 → 0 (R21 being the longest side Figure 3: Model of MPA above infinite ground plane for H =
of the triangle T) is satisfied, one needs only to use the first 10 mm, dominant mode TM01 shown.
term in the Taylor series expansion. The dominant singular
static part is 1/r21 and the integral to be worked out is 4.5

1 4
I= 2 dx d y dx dy , (16)
T 2
T (x − x ) + y − y 3.5 Regular
behaviour
f res (GHz)
where T = T is a triangular area. Using simplex coordinates 3

Min. f res
transformation (Figure 2), the result is [23, 24] 2.5
Min. Q
4 ln(1 − 2h12 /L) ln(1 − 2h13 /L) ln(1 − 2h23 /L) 2
I = − A2 + + ,
3 h12 h13 h23 1.5
(17)
1
where A is the triangle area, hi j are the edge lengths (see 0 5 10 15 20 25 30 35 40 45 50 55 60
Figure 2), and L is the perimeter of the triangle. H (mm)
f r (TCM)
4. Applications: Rectangular Patch Antenna f r (analytic equation)
Let us first concentrate on a rectangular patch antenna of Figure 4: R50 × 30 resonant frequency of the dominant TM01
dimensions L = 50 mm and W = 30 mm (further noted as mode. The dashed red curve is a quasianalytical equation from [1].
R50 × 30) placed in air at a height H above an infinite ground
plane. Only the dominant TM01 mode will be studied. The
reason for choosing a patch with L/W = / 1 is that we do not H/λres < 0.08), the resonant frequency decreases “regularly,”
have to deal with degenerated modes. and quasianalytical formulas (see, e.g., [1, 3]) based on
Using the image theory, the radiator in the XY plane the fringing field concept are valid below this range. For
at height z = H above an infinite electric ground plane is H ∼= 25 mm (H/λres ∼ = 0.188) there is absolute minimum
modelled as two patches separated by 2H. The total number of the TM01 resonant frequency. Further on, the resonant
of triangular elements is 676. In the TCM analyser, a proper frequency rises to reach its maximum for H ∼ = 40 mm
out-of-phase mode is selected (Figure 3). (H/λres ∼
= 0.51). Around this specific height the patch also
The resonant frequency of the dominant mode is shown shows the minimum of the radiation Q. The above described
as a function of height H, see Figure 4. It has been evaluated process repeats periodically. It is yet unclear to the authors
from a modal resonant condition for eigenvalues λ = as what is the physical background to the resonant frequency
2ω(W m − W e ) = 0 employing an adaptive frequency discontinuity around H/λres ∼= 0.5.
sweep for each height. The behaviour is quite peculiar, The terms 2ωW m , 2ωWe , and 2ω(W m − W e ) obtained
especially for greater heights. For low heights (H < 10 mm or from (5)–(9) and eigenvalues λ are plotted at Figure 5 for
20 an approximate analytical solution is available and in [25] we

showed that the Q is led by the function
15
2H
fQ (H) ∼
= . (18)
k2H − sin(k2H)
Reactive energies
10
f res After deriving (18), the condition is worked-out
5
tan(k2H) = k2H, (19)
0
and the first nontrivial root of (19) could be approximated
−5
as [25]

H ∼ 3 − 1 = 0.358.
−10
= 2
(20)
1.5 1.6 1.7 1.8 1.9 2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 λ min 8 6π
f (GHz) For sinusoidal currents on dipoles the minimum (evaluated
numerically) occurs for H = 0.36λ.
λ 2ω (We)
The minimum of the patch under study is obtained at
2ω (Wm) 2ω (Wm-We)
H ∼= 0.4λ, a value that is remarkably close to the simple
Figure 5: Reactive energies and their differences for an R50 × 30 dipole model.
patch at height of 25 mm.
4.1. Algorithm Convergence. Since no other methods for
calculating modal Q are available, Qeig is taken as a reference,
50
and the relative error percentage is defined as:
45

40 QJ − Qeig
relative error = · 100, (21)
35 Qeig
30
Q (–)
25 where QJ is calculated from the currents using (11). Four

different heights H were chosen, H = 1 mm (0.01λ), H =
20
2 mm (0.0185λ), H = 10 mm (0.0803λ), and H = 20 mm
15
(0.151λ), and the relative error was evaluated as a function of
10
total triangular elements (including the mirror), see Figure 7.
5 All quality factors were evaluated at the resonant frequency
0
0 5 10 15 20 25 30 35 40 45 50 55 60
of the dominant mode for the R50 × 30 patch. As discussed
H (mm)
earlier, the centroid approximation became more inaccurate
with low heights H. However, even for the lowest analyzed
Q eig value H = 0.01λ, the relative error is in the order of
QJ a few percent for reasonable mesh density (hundreds of
elements). Further improvements to the integration routine
Figure 6: The radiation Q for dominant mode of an R50 × 30 patch are considered for the future.
as a function of height H.
4.2. Fractional Bandwidth of the R50 × 30 Patch Antenna. It
is known that the fractional bandwidth (FBW) is related to
H = 25 mm as a function of frequency. There is excellent the unloaded Q factor and the desired matching VSWR level.
agreement between the difference in stored energies and the For VSWR < s we have [26]
eigenvalues, both obtained in a completely different manner.
s−1
There is also very good agreement between the exact QJ FBW ∼
= √ [%]. (22)
and Qeig confirming the validity of the proposed algorithm Q s
via (14), see Figure 6. Note that Qeig in (10) does not require Using a full-wave simulator CST-MWS [27], an R50 × 30
the currents to be calculated on the structure while QJ is patch has been simulated and the FBWCST for VSWR < 2 was
evaluated in a rigorous way from modal currents (11). calculated as:
From Figure 6 it is seen that the radiation Q has a f2 − f1
minimum for a specific height. It is deduced that the reason FBWCST = , (23)
f0
lies in the cancelling of the radiated power between the two
out-of-phase currents. Similar behaviour has been observed where f2 and f1 are margins for VSWR < 2 and f0 is
in the case of two half-wave thin-wire dipoles with opposite the centre frequency. Only very low heights were studied
sinusoidal currents, separated by d = 2H, see [25] for details. since we used a simple probe feed which introduces an
Actually these two out-of-phase dipoles may serve as a very inductance component to the total input impedance. The
simple model for a patch antenna with a dominant mode. comparison in Figure 8 shows good agreement of both
When the dipoles are reduced to elementary (Hertzian) ones, fractional bandwidths.
10
9 J1 J2
8
Relative error (%)
7
6
5
4
3
2
1
0 σ1 σ2
0 200 400 600 800 1000 1200 1400
Number of triangle elements
H = 1 mm (0.01 λ) H = 10 mm (0.0803 λ)
H = 2 mm (0.0185 λ) H = 20 mm (0.151 λ)
Figure 7: Relative error of the Q factor as a function of triangular

elements (mesh density). Figure 9: The first two characteristic modes (currents and charges)
for the SAU1 structure.
5
4.5
4 W
3.5
J1 J2
3
FBW (%)
2.5
L
2
1.5
1
0.5
0 σ1 σ2
1 1.5 2 2.5 3 3.5 4 4.5 5
H (mm)
FBW (CST)
FBW (Q eig )
0 0
Figure 8: Fractional bandwidth FBW (VSWR < 2) for a R50 × 30
patch. Figure 10: The first two characteristic modes (currents and
charges) for the SAU2 structure.
5. Applications: Fractal Antennas

In this section, a bit more complex structures will be studied.
The first one (the “Self Affine U” fractal, SAU), has been
described in [28] and further analyzed in [29]. This kind J1 J2
of radiating motif is employed as a dual-band radiator
Original
with mutually orthogonal radiation patterns at both bands.
Therefore we are analyzing the first two modes, where the
currents are orthogonal. These are depicted in Figures 9
and 10 for first (SAU1) and second (SAU2) fractal iteration, Infinite ground plane
respectively. The current of the first (lower) mode J1 has
Image
two out-of-phase components (see Figure 11 for schematic
current paths) while the second mode comprises inphase
currents only. As we know from previous studies, opposite
currents contribute to a rapid increase of the radiation Q, Figure 11: The main current paths for the first two modes of the
and it is expected that J1 will have a much higher Q than J2 . SAU1/2 structure.
280 100
260 90
240 80 Q1
220 70
Mode no.2 Mode no.3
200 60
αn
Mode no.1
Q
180 50
160 40
30
140 Q2
20
120
10
100
1 1.25 1.5 1.75 2 2.25 2.5 2.75 3 3.25 3.5 3.75 4 0
10 12 14 16 18 20 22 24 26 28 30
f (Hz) × 109
H (mm)
(a) (b)
Figure 12: Characteristic angles (left) and radiation Q for the SAU2.
0.03 0.03
0.02 J1 J2
0.02
0.01 0.01
0 0
−0.03 −0.02 −0.01 0 0.01 0.02 0.03 −0.03 −0.02 −0.01 0 0.01 0.02 0.03
(a) (b)
0.03 0.03
0.02 σ1 0.02 σ2
0.01 0.01
0 0
−0.01 −0.01
−0.02 −0.02
−0.03 −0.03
−0.03 −0.02 −0.01 0 0.01 0.02 0.03 −0.03 −0.02 −0.01 0 0.01 0.02 0.03
(c) (d)
Figure 13: Degenerated dominant mode J1 , and J2 of the FCL2 antenna (currents and charges).
Figure 11 presents a very simple concept showing the 5.1. The FCL-2 Fractal Antenna. The second presented struc-
main current paths for the J1 and J2 modes discussed above ture is the so-called fractal clover leaf (FCL) of the second
including the mirroring effect of the infinite ground plane. It iteration, [14]. The antenna is fed by an L-probe [30] that
could be simply stated that more opposing current paths lead excites its dominant mode and is located at height H =
to significant increase in Q. 36 mm. Actually, the dominant mode is composed of two
We show detailed behaviour only for SAU2 (the situation degenerated modes J1 and J2 (Figure 13). The second higher
is similar for SAU1)—see Figure 12 that confirms high Q for mode J3 is shown at Figure 14 for completeness.
the J1 mode. Characteristic angles are calculated for H =
29 mm, the actual height for which the dual-band antenna Figure 15 shows the main current paths of these modes,
was designed [29]. and we can again deduce that the dominant mode will exhibit
0.03 0.03
0.02 J3 0.02 σ3
0.01 0.01
0 0
−0.01
−0.02
−0.03
−0.03 −0.02 −0.01 0 0.01 0.02 0.03 −0.03 −0.02 −0.01 0 0.01 0.02 0.03
(a) (b)
Figure 14: Second higher mode J3 (currents and charges).
Mode 1, 2 8.25
J1 + J2 6.87
5.5
4.12
2.75
1.37
= 0
−3 .6 2
−7 .2 4
−1 0 .8 7
−1 4 .4 9
−1 8 .1 2
−2 1 .7 4
Mode 3 7.32
6.1
J3 4.88
3.66
2.44
1.22
= 0
−3 .7 7
−7 .5 5
−1 1 .3 3
−1 5 .1 1
−1 8 .8 9
−2 2 .6 7
Figure 15: Schematic depiction of the dominant current paths for the dominant (J1 + J2 ) and the second higher J3 modes together with their
modal radiation patterns.
lower Q compared to J3 . This is confirmed by Figure 16—J3 Secondly, it has been observed that resonant frequency is
has more than 200x higher radiation Q. quite a complicated function of height. Unfortunately we do
not yet have any physical explanation as to why some modes
present minimum values of fr .
6. Resonant Properties of Studied Antennas Looking at Figure 18, it is clear (and interesting) that
The properties of studied antennas are summarized in this the resonant frequency behaves quite differently for low-
section. At first we observed that microstrip antenna could Q and high-Q modes. The resonant frequency of low-Q
support different kinds of modes regarding their Q factors modes is much more sensitive to the height, whereas high-Q
(see Figure 17): modes exhibit almost constant fr when the height is varied.
The proposed explanation is that the opposite currents
(a) low Q modes with the current flowing in one direc- (responsible for high Q) keep reactive fields very close to the
tion and not changing its phase (dominant modes of radiating structure so the effect of a fringing field coupled to
simple shapes like rectangular, circular patch, and so the ground plane becomes almost negligible.
forth.)
(b) high Q modes with part of the currents flowing in the 7. Conclusions
opposite direction. These modes exist even on simple
“U” shaped patch (Figure 9 left) and on complex Modal resonant properties of selected microstrip patch
(fractal) geometries. antennas have been studied with the help of characteristic
280 3
260 2.8
Low-Q
240 2.6 modes
220 2.4
2.2
fr (GHz)
200 Qeig3 = 226.3
2
αn
180
High-Q
1.8
160 modes
Qeig1 = Qeig2 = 10.5 1.6
140
1.4
120
1.2
100
1
80 0 5 10 15 20 25 30 35
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 2 H (mm)
f (GHz) × 109
FCL2 SAU2 mode 2
Mode 1 SAU1 mode 1 R50 × 50
Mode 2 SAU1 mode 2 R50 × 30
Mode 3 SAU2 mode 1
Figure 16: Characteristic angles for the FCL2 structure at H = Figure 18: Resonant frequencies for different antennas/modes.
29 mm.
100 Using proper feeding techniques (like the L-probe) allows us

High-Q to design wideband compact antennas.
90
modes The theory now puts current distribution and the radi-
80
ation Q factor into objective context. Whenever the current
70
mode exhibits opposite components, high Q may appear.
60 Future work is needed to connect the presented theory
Q eig
50 with parameter sweeps or even optimization, so we will be

40 Low-Q able to design novel wideband/multimode compact anten-
modes nas.
30
20
10 Acknowledgments
0
0 5 10 15 20 25 30 35 This work was supported by the Grant Agency of the
H (mm) Czech Technical University in Prague, grant no. SGS11/065/
OHK3/1T/13 and by the Project COST 1102. The authors
FCL2 SAU2 mode 2
SAU1 mode 1 R50 × 50
would like to thank professor Vandenbosch for fruitful
SAU1 mode 2 R50 × 30 discussions, Neil Bell for his comments, and two anonymous
SAU2 mode 1 reviewers who suggested some improvements to the paper.
Figure 17: Radiation Qs for different antennas/modes.
References
[1] P. Bhartia, I. Bahl, R. Garg, and A. Ittipiboon, Microstrip
Antenna Design Handbook, Artech House, Norwood, Mass,
modes and the novel theory published by Vandenbosch. It
USA, 2000.
has been found that the resonant frequency of a simple
[2] K. F. Lee and W. Chen, Advances in Microstrip and Printed
rectangular patch antenna is quite a complicated function
Antennas, John Wiley & Sons, New York, NY, USA, 1997.
of its height above the infinite ground. Moreover, the
[3] J. R. James, P. S. Hall, and C. Wood, Microstrip Antenna Theory
dependency of resonant frequency is also found to be a and Design, Peter Peregrinus, London, UK, 1986.
function of the radiation Q factor (which is now possible to [4] R. F. Harrington and J. Mautz, “Theory of characteristic
calculate in a rigorous way). Due to the complexity of the modes for conducting bodies,” IEEE Transactions on Antennas
problem, no physical explanation for the resonant frequency and Propagation, vol. 19, no. 5, pp. 622–628, 1971.
behaviour has yet been found. [5] E. A. Daviu, Analysis and design of antennas for wireless
It is observed that the radiation Q factor decreases communications using modal methods, Ph.D. dissertation,
for “standard” heights (<∼0.1λ), however there exists an Universidad Politécnica de Valencia, Valencia, Spain, 2008.
absolute minimum value of Q that has already been predicted [6] G. A. E. Vandenbosch, “Reactive energies, impedance, and Q
by simple modeling of two elementary out-of-phase dipoles. factor of radiating structures,” IEEE Transactions on Antennas
and Propagation, vol. 58, no. 4, Article ID 5398856, pp. 1112– [28] S. N. Sinha and M. Jain, “A self-affine fractal multiband
1127, 2010. antenna,” IEEE Antennas and Wireless Propagation Letters, vol.
[7] G. A. E. Vandenbosch and V. Volski, “Lower bounds for 6, Article ID 891519, pp. 110–112, 2007.
radiation Q of very small antennas of arbitrary topology,” in [29] J. Eichler, P. Hazdra, M. Čapek, T. Kořı́nek, and P. Hamouz,
Proceedings of the 4th European Conference on Antennas and “Design of a dual-band orthogonally polarized l-probe-fed
Propagation (EuCAP ’10), pp. 1–4, Barcelona, Spain, April fractal patch antenna using modal methods,” IEEE Antennas
2010. and Wireless Propagation Letters, vol. 10, pp. 1389–1392, 2011.
[8] R. F. Harrington, Field Calculation by the Method of Moments, [30] C. L. Mak, K. M. Luk, K. F. Lee, and Y. L. Chow, “Experimental
IEEE Press, New York, NY, USA, 1993. study of a microstrip patch antenna with an L-shaped probe,”
[9] M. Cabedo-Fabres, E. Antonino-Daviu, A. Valero-Nogueira, IEEE Transactions on Antennas and Propagation, vol. 48, no. 5,
and M. F. Bataller, “The theory of characteristic modes pp. 777–783, 2000.
revisited: a contribution to the design of antennas for modern
applications,” IEEE Antennas and Propagation Magazine, vol.
49, no. 5, pp. 52–68, 2007.
[10] P. Hazdra and P. Hamouz, “On the modal superposition lying
under the MoM matrix equations,” Radioengineering, vol. 17,
no. 3, pp. 42–46, 2008.
[11] Mathworks, http://www.mathworks.com/.
[12] S. N. Makarov, Antenna and EM Modeling with Matlab, John
Wiley & Sons, New York, NY, USA, 2002.
[13] S. M. Rao, D. R. Wilton, and A. W. Glisson, “Electromagnetic
scattering by surfaces of arbitrary shape ,” IEEE Transactions on
Antennas and Propagation, vol. 30, no. 3, pp. 409–418, 1982.
[14] M. Capek, P. Hazdra, P. Hamouz, and M. Mazanek, “Software
tools for efficient generation, modelling and optimisation of
fractal radiating structures,” IET Microwaves, Antennas and
Propagation, vol. 5, no. 8, pp. 1002–1007, 2011.
[15] Comsol Multiphysics, http://www.comsol.com/.
[16] J. L. Volakis, Ch. Chen, and K. Fujimoto, Small Anten-
nas: Miniaturization Techniques & Applications, chapter 2,
McGraw-Hill, New York, NY, USA, 1st edition, 2010.
[17] J. S. McLean, “A re-examination of the fundamental limits
on the radiation q of electrically small antennas,” IEEE
Transactions on Antennas and Propagation, vol. 44, no. 5, pp.
672–676, 1996.
[18] R. F. Harrington and J. R. Mautz, “Control of radar scattering
by reactive loading,” IEEE Transactions on Antennas and
Propagation, vol. 20, no. 4, pp. 446–454, 1972.
[19] R. F. Harrington, Time-Harmonic Electromagnetic Fields, IEEE
Press, New York, NY, USA, 2001.
[20] D. R. Rhodes, “Observable stored energies of electromagnetic
systems,” Journal of the Franklin Institute, vol. 302, no. 3, pp.
225–237, 1976.
[21] D. R. Rhodes, “Reactance Theorem,” Proceedings of the Royal
Society A, vol. 353, no. 1672, pp. 1–10, 1977.
[22] J. Shaeffer, “MOM3D method of moments code theory
manual,” Research report 189594, Lockheed Advanced Devel-
opment Company, 1992.
[23] M. Capek, P. Hazdra, and J. Eichler, “A method for the
evaluation of radiation Q based on modal approach,” IEEE
Transactions on Antennas and Propagation. In press.
[24] P. Arcioni, M. Bressan, and L. Perregrini, “On the evaluation
of the double surface integrals arising in the application
of the boundary integral method to 3-D problems,” IEEE
Transactions on Microwave Theory and Techniques, vol. 45, no.
3, pp. 436–439, 1997.
[25] P. Hazdra, M. Capek, and J. Eichler, “Radiation Q-factors of
thin-wire dipole arrangements,” IEEE Antennas and Wireless
Propagation Letters, vol. 10, pp. 556–560, 2011.
[26] A. D. Yaghjian and S. R. Best, “Impedance, bandwidth, and Q
of antennas,” IEEE Transactions on Antennas and Propagation,
vol. 53, no. 4, pp. 1298–1324, 2005.
[27] CST Gmbh, http://www.cst.com/.
doi:10.1155/2012/167658
Research Article
Tunable Compact UHF RFID Metal Tag Based on
CPW Open Stub Feed PIFA Antenna
Lingfei Mo and Chunfang Qin

State Key Laboratory of Industrial Control Technology, Department of Control Science and Engineering, Zhejiang University,
Hangzhou 310027, China
Correspondence should be addressed to Lingfei Mo, lfmo@iipc.zju.edu.cn
Received 15 August 2011; Revised 9 December 2011; Accepted 28 December 2011
Academic Editor: Seong-Youp Suh
Copyright © 2012 L. Mo and C. Qin. This is an open access article distributed under the Creative Commons Attribution License,
For the ultrahigh frequency radio frequency identification (UHF RFID) metal tag, it always has the difficulties of compact
designing, especially for the conjugate impedance matching, low antenna gain, and fabrication or environmental detuning. In
this paper, a tunable compact UHF RFID metal tag is designed based on CPW open stub feed PIFA antenna. By changing the
length of the open stub, the impedance of the PIFA antenna could be tuned in a large scale for conjugate impedance matching. The
open stub makes it easy to tune the resonant frequency to alleviate the fabrication detuning or the environmental detuning, even
after the manufacture. Moreover, the CPW structure of the open stub feed can resist the effects of the metallic surface and increase
the antenna gain for the compact PIFA antenna. Modeling analysis and simulation are in good agreement with the measurement
results. It showed that the UHF RFID metal tag could be designed compact with good performance based on the CPW open stub
feed PIFA antenna.
1. Introduction because the metallic surface boundary changes the radiation

efficiency, impedance matching, resonant frequency, and
Ultra high frequency radio frequency identification (UHF radiation pattern [2, 3]. For this reason, some special tags
RFID) is a long-range noncontact automatic identification which can be applied to the metallic surface were designed,
technology being widely used around the world recently. A called the metal mountable tags or metal tags. Impedance
UHF RFID system is generally composed of a reader and matching is very important for the UHF RFID metal tag
a tag. The reader reads the information of the tag through antenna design. As the input impedance of the UHF RFID
radio frequency (RF) wave. The tag is composed of a chip tag chip is generally complex, the antenna impedance should
and an antenna, with no internal battery. All the energy it be designed to be the conjugate impedance of the chip for
needs is obtained from the RF wave transmitted by the reader impedance matching [1]. Therefore, the input impedance
[1]. Compared with the traditional bar code tag, the UHF of the antenna should be adjusted flexibly to achieve good
RFID tag can be read and written over a long distance with conjugate impedance matching. Microstrip antenna, patch
a very high data rate, no matter whether the tag is soiled antenna, and PIFA antenna, which can take the metallic
or dirty. So, UHF RFID technology has been adopted in surface as ground plane, are designed for RFID metal tags
the logistics supply, automatic manufacture, traffic manage- with different impedance matching methods [4–8].
ment, property security, and so forth. In some applications, Compact UHF RFID metal tag is widely used in practical
the UHF RFID tags need to be attached on the surface of applications. However, for compact antenna design, the
metallic objects, such as steel plates and steel containers. impedance matching of these antennas is limited because
However, when UHF RFID tags, especially those with dipole- of the limited antenna structure. The impedance of the
like antennas, are placed on the metallic surface, the reading compact antenna should be able to be tuned flexibly in a large
distance is reduced quickly, or even cannot be read. This is adjusting scale to match different tag chips. Furthermore,
for commercial UHF RFID tags manufacture, the impedance Power + query
Antenna Chip
may not be matched well between the antenna and the chip
because of the simulation error and the fabrication variabil- Tag
ity, such as the substrate permittivity difference, manufactur- Reader Za Zc
ing difference, and chip impedance difference. To get good
impedance matching and performance, some manufacturers
use automatic laser or mill machine to adjust the antenna Backscattered wave
physical structure to tune the impedance of the antenna.
For this purpose, the tag antenna should be designed to Antenna impedance Chip impedance
have an easy tuning structure, which is easy for laser milling Figure 1: Principle of back-scattered UHF RFID system.
machine to adjust and achieving good impedance matching.
Impedance tuning is also very useful to alleviate the detuning
effects due to the different metallic application environments
For UHF RFID tag, one of the most important criteria of
[2]. Besides the impedance matching and tuning, antenna performance is the read range. The maximum read range of
gain is another challenge for the compact RFID metal tag. the tag can be calculated as [18]
Compact RFID metal tags always have lower antenna gain

and shorter read range. PIFA antenna is widely used for λ Pt Gt Gr τ
the compact UHF RFID tag mountable on metallic objects r= , (1)
4π Pth
[9–11]. However, with radiation patch size reduction, the
antenna gain decreases and the impedance matching become where λ is the free space wavelength, Pt is the power
difficult. Using two PIFAs can get better antenna gains [12, transmitted by the reader, Gt is the gain of the antenna of
13] and flexible impedance matching [5, 14]. But the size the reader, Gr is the gain of the antenna of the tag, τ is the
would increase obviously. Therefore, all of these previous power transmission coefficient between the tag antenna and
works cannot fit the requirement of compact profile, easy the chip, and Pth is the threshold power of the chip. When the
tuning, and satisfied antenna gain. reader and the chip of the tag keep the same, the maximum
In this paper, a tunable compact UHF RFID metal tag read range of the UHF RFID tag is mainly determined by the
antenna with a CPW open stub feed is proposed. Through design of the tag antenna, especially the gain of the antenna
PIFA antenna, the tag was compact designed. The impedance of the tag (Gr ) and the power transmission coefficient (τ)
of the antenna could be tuned freely by changing the [2]. The power transmission coefficient τ is determined by
length of the open stub [15]. Because of the open stub the impedance matching of the chip and the antenna, which
design, the impedance matching can be tuned even after can be calculated as follows:
the fabrication of the tag. Together with the open stub, the 4Rc Ra
CPW is used to resist the effects of the metallic objects τ= , 0 ≤ τ ≤ 1, (2)
| Z c + Za | 2
[16] and improve the performance of antenna gain [17]. In
Section 2, the considerations of the UHF RFID tag design are where Zc = Rc + jXc is the impedance of the tag chip,
discussed first. Based on these considerations, the proposed Za = Ra + jXa is the impedance of the tag antenna. When
antenna is described in Section 3, with antenna structure, the impedances of the antenna and the chip are conjugate
theoretical modeling and simulation results. In Section 4, matching, the transmission coefficient τ could get the maxi-
measurement results of a prototype based on this design mum value 1 and the most energy will be transmitted from
are also provided. Finally, discussion and conclusions are the antenna to the chip when the tag is being enquired by the
presented in Section 5. reader.
Besides the gain and the impedance matching, the band-
width and the radiation pattern are also important consid-
2. UHF RFID Tag Design Considerations erations for UHF RFID tag antenna design. Wide bandwidth
makes the tag to be read in a required bandwidth and the
A typical passive RFID tag is composed of a chip and an broadside radiation pattern makes the tag to be read in a
antenna, with no internal battery. All the energy it needs wide direction scale [19, 20]. In the realistic application, the
is obtained from the electromagnetic wave transmitted by a size and shape of the tag must be designed to be embedded or
RFID reader. In a passive back-scattered UHF RFID system, attached to the target objects and have a reliable performance
the reader transmits a modulated signal with periods of un- [21–23]. And for a commercial RFID tag, the cost is also
modulated carry wave, which is received by the antenna an important requirement to be considered [24, 25]. The
of the tag. When the chip of the tag is activated by the cost of the RFID tag is a critical factor for this technology
power from the antenna, it will send back its identification to be widely used around the world. Another requirement
information by modulating the backscattered signal. The for RFID tag antenna design is the easiness for the mass
backscattered signal is modulated by switching the load production [26]. This includes the antenna manufacture,
impedance of the chip between two states [1]. Figure 1 the chip bonding, the tag package, the performance testing,
illustrates the operation of back-scattered passive UHF RFID and the frequency tuning. The frequency tuning is useful
system. for reducing the differences of chips and substrate materials,
keeping the tags with the same performance before going to Table 1: Requirements for designing a UHF RFID tag antenna.
be used.
Requirements of UHF RFID Effects of improvement
The requirements for designing a UHF RFID tag antenna
tag antenna design Read range Cost Reliability
are concluded in Table 1. The proposed tunable compact tag
√
antenna is designed according to these requirements. (1) Good impedance matching
√
(2) High antenna gain
3. Proposed Antenna Design √ √
(3) Wide bandwidth
√ √
A tunable compact UHF RFID metal tag antenna is proposed (4) Broadside radiation pattern
√
in this paper. The PIFA antenna design makes the antenna (5) Compact size and shape
√
compact than normal microstrip antenna. With an open stub (6) Low manufacture cost
√ √ √
feed, the antenna can be conjugate impedance matched with (7) Easy for testing and tuning
the chip easily by tuning the inset depth and the open stub √ √
(8) Stable performance for use
length [15]. This metal tag antenna can be fabricated cheaply
with normal PCB (Printed Circuit Board) technology. More- CPW open stub Shorting wall
over, with the open stub as tuning structure, the working h Chip
frequency of the tag can be tuned even after the tag has been
manufactured. The antenna bandwidth, radiation pattern,
and metal stability also keep with good performance. Ls Radiation patch
Linset
3.1. Antenna Structure. The structure and dimensions of the Winset
proposed antenna are illustrated in Figure 2. It is a planar L
W
inverted F antenna with a shorting wall to reduce size. The Z
radiation patch has dimensions of W (20 mm) × L (38 mm) Y
and is printed on a FR4 substrate (εr = 4.4, tan δ = 0.02).
The dimension of the substrate is (W + 2 mm) × (L + X Ground
2 mm) × h (3 mm). The open stub feed line is inset into the
patch to decrease the input impedance of the patch [27]. The
inset structure has a length of Linset and a width of Winset Figure 2: The open stub feed PIFA antenna structure. The dimen-
(8 mm). The open stub feed line has a length of Ls and a sions of the radiation patch are (L, W) while the open stub has
width of Ws (3 mm). The chip is attached on the feed port dimensions (Ls , Ws ) and the height of the substrate (FR4) between
the patch and the metallic surface is h.
composed by the open stub line and the radiation patch. In
order to enhance the gain of the compact PIFA antenna, a
CPW structure is designed for the open stub feed line. The
antenna is attached on a 200 mm × 200 mm metal plate. The The input impedance of the open stub only has imagi-
parameters Linset and Ls are used as variables for impedance nary part and its function of line length is shown in Figure 4.
matching. It shows that the reactance of the CPW open stub feed line is
capacitive when the length is less than 0.25 wavelength and
3.2. Theoretical Modeling Analysis. The transmission line is inductive when the length is between the 0.25 wavelength
model of the antenna is shown in Figure 3. From the antenna and 0.5 wavelength. The reactance of the CPW open stub is
model, it is easy to know that the radiation patch and the a function of cotangent, which means that when the length
CPW open stub feed lines are in series. Therefore, the input of the open stub changes from 0 to 0.5 λ, the imaginary part
impedance of the feed port of the antenna can be calculated of the input impedance of the open stub changes from −∞
as to +∞. Therefore, the imaginary part of the input impedance
of the antenna can be tuned freely by the length of the open
1
Zin = Zin 2
+ Zin , (3) stub in a large scale.
1
where Zin is the input impedance of the radiation patch of the 3.3. Simulation and Optimization. In order to get a bet-
PIFA antenna, Zin 2
is the input impedance of the CPW open ter impedance matching for the antenna, Finite-Element-
stub feed line. Method (FEM) based computational simulation software
According to the basic RF circuit theory [28], the input HFSS 12 is used for the simulation and optimization. For
impedance of the open stub can be simplified as the UHF RFID tags, the chips generally have complex
impedance, whose imaginary part is large and negative
1 1 because of the rectifier and energy storage capacitor. In
2
Zin = − jZ02 = − jZ02 , (4)
tan βLs tan(2πLs /λ) order to achieve the maximum energy transfer between the
antenna and the chip, the input impedance of the antenna
where Z02 is the characteristic impedance of the CPW open and the chip should be conjugate matching. That is, the
stub feed line, β is the wave number, Ls is the length of the real part is equal, and the imaginary part is opposite. As
CPW open stub feed line. the imaginary part is much larger than the real part of the
CPW open stub 2

Zin 1
Zin PIFA patch be used to tune the imaginary part of the input impedance
of the antenna, which is shown in Figure 5(b). The reactance
Feed of the antenna increases with the increases of the CPW open
Ls , β stub length Ls . The imaginary part of the input impedance
ZL2 = ∞ Z02 1
Zin of the antenna could be tuned freely from −∞ to +∞ by
changing the length of the CPW open stub from 0 to 0.5 λ.
Therefore, for conjugate impedance matching of the
proposal antenna, the resistance and the reactance could
Figure 3: Transmission line model of the CPW open stub feed PIFA be tuned freely by the depth of the inset (Linset ) and the
antenna. length of the CPW open stub (Ls ), respectively. Through
simulation and optimization, the parameters of the antenna
are finalized as L = 37 mm, W = 20 mm, h = 3 mm,
5 Linset = 14 mm and Ls = 30 mm. With this dimension, the
antenna input impedance and the reflection coefficient S11
are calculated as shown in Figures 6(a) and 6(b), respectively.
2.5 The imaginary parts of the impedance of the antenna and
Inductive the chip are matched well at the frequency of 914 MHz.
And the real parts of the impedance are matched at the
/( jZ0 )
0
frequency of 918 MHz. However, as the imaginary part of the
impedance is much larger than the real part, the impedance
Zin
2
matching is dominated by the imaginary part. Under this

Capacitive matching condition, the reflection coefficient S11 is located
−2.5 at the 914 MHz with a value of −26 dB. The 3 dB bandwidth
of the antenna is 37 MHz (895 MHz∼932 MHz), covering the
bandwidth of the UHF RFID of North American. Moreover,
−5 with the decreases of the length of the open stub, the working
0 0.1 0.2 0.3 0.4 0.5 frequency of the antenna is tuned from low to high.
L/λ
Figure 4: The input impedance of the CPW open stub feed line.
4. Measurement Results
Based on the above-optimized parameters, the antenna
sample was produced with an FR4 dielectric plate, as shown
impedance, the impedance matching is mainly determined in Figure 7. The chip was attached to the antenna feed port
by the imaginary part matching. So, the antenna should with the traditional bonding technology. In order to test and
be designed to have a structure easy for impedance tuning, compare the performance with the simulation results, the
especially for imaginary part tuning. As the proposed tag tag was also mounted on a 200 × 200 mm2 copper plate.
antenna in this paper is designed for the North America UHF A commercial RFID reader, CSL-461 4-Port EPC Class1
RFID bandwidth (902 MHz∼928 MHz), the tag antenna Gen2 UHF RFID Reader [29], was used to test the tag. The
should have good impedance matching at this bandwidth. bandwidth of the reader is 902 MHz∼928 MHz. The output
The chip used for the tag is the RI UHF 00001 01 UHF power of the reader can be tuned from 15 dBm to 30 dBm,
RFID chip of TI (Texas Instruments), whose impedance is with a step of 0.25 dBm. The antenna of the reader is CS-771-
9.9-j60.3 Ω at the frequency of 915 MHz. The structure of 2-R with a gain of 6 dBi. Combining the output power of the
the antenna is shown in Figure 2. In order to simulate the reader and the reader antenna gain, the maximum radiation
tag antenna on the surface of metal, the tag is simulated on power is 36 dBm (4 W EIRP). The reader and the tag are
the surface of a reference metallic plate of 200 × 200 mm2 . manufactured with the protocol of EPC Class1 Gen2 and
According to the relative permittivity of the substrate, the ISO 18000-6C. According to the tag performance parameters
length of the radiation patch (L) of the PIFA antenna is and test methods of EPCglobal, the performance of the tag
chosen as 37 mm, which makes the PIFA antenna resonant was measured based on the back-scattering method [30].
near the frequency of 915 MHz. The impedance matching The maximum read range, power bandwidth, and radiation
between the antenna and the chip is tuned by Linset and pattern were measured with the same method.
Ls . Through the back-scattering method, the best impedance
For patch antennas with the inset feed structure, increas- matching frequency (resonant frequency) of the antenna
ing the depth of the inset could decrease the input impedance could be measured. Because of the fabrication variability,
of the antenna [27]. Therefore, the length of the inset Linset the best impedance matching frequency is a little lower than
can be used to tune the real part of the antenna impedance. 915 MHz. Then, as shown in Figures 5(b) and 6(b), through
Figure 5(a) shows the resistance tuning of the proposed cutting the length of the open stub, the imaginary part of the
antenna with different inset depths (Linset ). The resistance of impedance can be decreased and the resonant frequency can
the antenna decreases with the increase of the inset depth be increased to the target working frequency. In this way, the
Linset . As we analyze above, the CPW open stub feed line can fabricated tag prototype is optimized by tuning the length
20 150
120
15
Resistance (ohm)
Reactance (ohm)
90
10
60
Chip resistance
5
30 Chip reactance
conjugate
0 0
0.86 0.88 0.9 0.92 0.94 0.96 0.86 0.88 0.9 0.92 0.94 0.96
Frequency (GHz) Frequency (GHz)
Linset = 12 mm Linset = 15 mm Ls = 28 mm Ls = 31 mm
Linset = 13 mm Linset = 16 mm Ls = 29 mm Ls = 32 mm
Linset = 14 mm Ls = 30 mm
(a) (b)
Figure 5: Antenna input impedance tuning. (a) Input resistance curves of the antenna with different inset depths Linset (L = 37 mm, W =
20 mm, h = 3 mm, Ls = 30 mm). (b) Input reactance curves of the antenna with different open stub length Ls (L = 37 mm, W = 20 mm,
h = 3 mm, Linset = 14 mm).
100 0
37 MHz
−5
80 895 MHz 932 MHz
Reflection coefficient S11 (dB)
Chip reactance conjugate −10

Antenna reactance
Impedance (ohm)
60 −15
Reactance matching point −20

40
−25
Resistance matching point Antenna resistance

20 −30 914 MHz
Chip resistance
−35
0
0.86 0.88 0.9 0.92 0.94 0.96 −40
0.86 0.88 0.9 0.92 0.94 0.96
Frequency (GHz)
Frequency (GHz)
Antenna impedance
Chip conjugate impedance Ls = 28 mm Ls = 31 mm
Ls = 29 mm Ls = 32 mm
Ls = 30 mm
(a) (b)
Figure 6: The input impedance and reflection coefficient S11 of the antenna with optimized parameters: (a) impedance, (b) reflection
coefficient S11 .
of the open stub to alleviate the detuning effects due to the same. The power bandwidth of the tag was measured in
fabrication process. Figure 9. The output power of the reader needed to read
The maximum read range of the tag in North America the tags at different frequencies was normalized with the
bandwidth is plotted in the Figure 8. The antenna has a stable minimum value. The minimum is 0 dB. From Figure 9, it is
read range in the whole North America UHF RFID band easy to calculate that the 3 dB power bandwidth is 903 MHz∼
with a max value of 4.7 meters at the frequency of 915 MHz. 927 MHz, which covers most of the North America UHF
The tested results and the theoretical values are almost the RFID bandwidth. Compared with the 3 dB bandwidth of the
Power of the reader (dB)

4
24 MHz
3
1
Figure 7: Photograph of the fabricated tag antenna.
0
904 908 912 916 920 924 928
6 Frequency (GHz)
Figure 9: Power bandwidth of the tag antenna.

5
4 Table 2: Tag performance testing results with the tag attached on

Red range (m)
different metal plate.

3
Size of the metal plate Read ranges
200 mm × 200 mm 4.7 m
2
500 mm × 500 mm 4.8 m
1
700 mm × 700 mm 4.3 m
0
904 908 912 916 920 924 928 of the open stub even after the tag has been fabricated.
Frequency (GHz) This can be used to alleviate the detuning effects of the
fabrication error and the metallic application environments.
Testing results
Simulation results
With deceasing the length of the open stub, the imaginary
part of the antenna can be reduced and the working
Figure 8: Theoretical and experimental read ranges for the open frequency can be increased. With increasing the length of the
stub feed patch antenna (EIRP = 4 W). open stub, the imaginary part of the antenna can be increased
and the working frequency can be decreased.
The testing results were in good agreement with the
simulation. This antenna has stable performance on different
reflection coefficient S11 , the 3 dB power bandwidth is a little
sizes of metallic objects. Four features can be concluded for
narrow, but the central bandwidth is almost the same. The
this antenna design as follows.
measured radiation patterns of the tag at the frequency of
915 MHz are shown in Figure 10. The antenna has nearly
broadside hemisphere radiation pattern performance at both (1) By PIFA antenna design, the size of the tag can be
E plane and H plane. Tested results agree well with the effectively reduced. The length of the PIFA antenna
simulation results. The tag was attached on the metallic is only one half of that of microstrip antenna.
plates of different size to test its metal performance. The (2) An open stub feed is used to realize the impedance
testing results are plotted in Table 2, which shows that the matching for this compact PIFA antenna. The impe-
metal tag has stable read range when it is placed on the dance matching between the antenna and the chip
surface of different metallic objects. could be achieved easily by tuning the length of
the CPW open stub feed line. And this impedance
5. Discussion and Conclusion matching method could be used with different chips
and input impedances.
A tunable compact UHF RFID metal tag based on CPW open
stub feed PIFA antenna is designed in this paper. Using CPW (3) With the CPW open stub feed line, the impedance
open stub feed line, the impedance matching and antenna matching of the tag could be tuned even after the
gain can be well designed. Moreover, because of the PIFA manufacture of the tag. This makes it suitable for
and the CPW structure, the antenna has stable performance accurate impedance matching of the UHF RFID tag
for attaching on the surface of metallic objects. The working for manufacture and different application environ-
frequency of the antenna can be tuned by milling the length ments.
0 0
0 0
330 30 330 30
−10 −10
−20 300 60 −20 300 60
−30 −30
(dB)
(dB)
−40 270 90 −40 270 90
−30 −30
−20 −20
240 120 240 120
−10 −10
210 150 210 150

0 0
180 180
Simulation Simulation
Testing Testing
(a) (b)
Figure 10: Radiation patterns of the open stub feed PIFA antenna: (a) E plane, (b) H plane.
(4) The CPW structure of the open stub feed can [7] K. H. Kim, J. G. Song, D. H. Kim, H. S. Hu, and J. H.
resist the effects of the metallic surface and increase Park, “Fork-shaped RFID tag antenna mountable on metallic
the antenna gain for the PIFA antenna, which will surfaces,” Electronics Letters, vol. 43, no. 25, pp. 1400–1402,
keep the metal tag having a stable performance for 2007.
attaching on the surface of different metallic objects. [8] H.-W. Son and S.-H. Jeong, “Wideband RFID tag antenna
for metallic surfaces using proximity-coupled feed,” IEEE
Antennas and Wireless Propagation Letters, vol. 10, pp. 377–
Acknowledgment 380, 2011.
[9] M. Hirvonen, P. Pursula, K. Jaakkola, and K. Laukkanen,
This work was supported by the major projects of the Edu- “Planar inverted-F antenna for radio frequency identification,”
cation Administration (Y200907699), Zhejiang province, Electronics Letters, vol. 40, no. 14, pp. 848–850, 2004.
China. [10] H. Kwon and B. Lee, “Compact slotted planar inverted-F RFID
tag mountable on metallic objects,” Electronics Letters, vol. 41,
References no. 24, pp. 1308–1310, 2005.
[11] W. Choi, H. W. Son, J. H. Bae, G. Y. Choi, C. S. Pyo, and
[1] K. Finkenzeller, RFID Handbook, John Wiley & Sons, New J. S. Chae, “An RFID tag using a planar inverted-f antenna
York, NY, USA, 2nd edition, 2003. capable of being stuck to metallic objects,” Electronics and
[2] L. F. Mo, H. J. Zhang, and H. L. Zhou, “Analysis of dipole- Telecommunications Research Institute Journal, vol. 28, no. 2,
like ultra high frequency RFID tags close to metallic surfaces,” pp. 216–218, 2006.
Journal of Zhejiang University A, vol. 10, no. 8, pp. 1217–1222, [12] J. S. Kim, W. Choi, and G. Y. Choi, “UHF RFID tag antenna
2009. using two PIFAs embedded in metallic objects,” Electronics
[3] K. Penttilä, M. Keskilammi, L. Sydänheimo, and M. Kivikoski, Letters, vol. 44, no. 20, pp. 1181–1182, 2008.
“Radio frequency technology for automated manufacturing
[13] S. L. Chen and K. H. Lin, “A slim RFID tag antenna design
and logistics control. Part 2: RFID antenna utilisation in
for metallic object applications,” IEEE Antennas and Wireless
industrial applications,” International Journal of Advanced
Propagation Letters, vol. 7, Article ID 2009473, pp. 729–732,
Manufacturing Technology, vol. 31, no. 1-2, pp. 116–124, 2006.
2008.
[4] S. J. Kim, B. Yu, Y. S. Chung, F. J. Harackiewicz, and B.
Lee, “Patch-type radio frequency identification tag antenna [14] L. Xu, B. J. Hu, and J. Wang, “UHF RFID tag antenna with
mountable on metallic platforms,” Microwave and Optical broadband characteristic,” Electronics Letters, vol. 44, no. 2, pp.
Technology Letters, vol. 48, no. 12, pp. 2446–2448, 2006. 79–81, 2008.
[5] B. Yu, S. J. Kim, B. Jung, F. J. Harackiewicz, and B. Lee, “RFID [15] L. Mo and C. Qin, “Planar UHF RFID tag antenna with open
TAG antenna using two-shorted microstrip patches mount- stub feed for metallic objects,” IEEE Transactions on Antennas
able on metallic objects,” Microwave and Optical Technology and Propagation, vol. 58, no. 9, Article ID 5484680, pp. 3037–
Letters, vol. 49, no. 2, pp. 414–416, 2007. 3043, 2010.
[6] B. Lee and B. Yu, “Compact structure of UHF band RFID [16] C. H. Ku, H. W. Liu, and P. J. Wang, “Novel CPW-fed slot
tag antenna mountable on metallic objects,” Microwave and antenna for UHF RFID metal tag applications,” IEICE Elec-
Optical Technology Letters, vol. 50, no. 1, pp. 232–234, 2008. tronics Express, vol. 8, pp. 410–415, 2011.
[17] Y. Um, U. Kim, and J. Choi, “Design of a compact CPW-fed

UHF RFID tag antenna for metallic objects,” Microwave and
Optical Technology Letters, vol. 50, no. 5, pp. 1439–1443, 2008.
[18] K. V. S. Rao, P. V. Nikitin, and S. F. Lam, “Antenna design for
UHF RFID tags: a review and a practical application,” IEEE
3870–3876, 2005.
[19] L. Mo, H. Zhang, and H. Zhou, “Broadband UHF RFID tag
antenna with a pair of U slots mountable on metallic objects,”
Electronics Letters, vol. 44, no. 20, pp. 1173–1174, 2008.
[20] M. Lai and R. Li, “Broadband UHF RFID tag antenna with
parasitic patches for metallic objects,” Microwave and Optical
Technology Letters, vol. 53, no. 7, pp. 1467–1470, 2011.
[21] D. Kim and J. Yeo, “A passive RFID tag antenna installed in a
recessed cavity in a metallic platform,” IEEE Transactions on
Antennas and Propagation, vol. 58, no. 12, Article ID 5582265,
pp. 3814–3820, 2010.
[22] H. D. Chen and Y. H. Tsao, “Low-profile meandered patch
antennas for RFID tags mountable on metallic objects,” IEEE
Antennas and Wireless Propagation Letters, vol. 9, Article ID
5419020, pp. 118–121, 2010.
[23] K. Ide, S. Ijiguchi, and T. Fukusako, “Gain enhancement of
low-profile, electrically small capacitive feed antennas using
stacked meander lines,” International Journal of Antennas and
Propagation, vol. 2010, Article ID 606717, 8 pages, 2010.
[24] S. Merilampi, L. Ukkonen, L. Sydanheimo, P. Ruuskanen, and
M. Kivikoski, “Analysis of silver Ink bow-Tie RFID tag anten-
nas printed on paper substrates,” International Journal of
Antennas and Propagation, vol. 2007, Article ID 90762, 9 pages,
2007.
[25] L. Yang, A. Rida, R. Vyas, and M. M. Tentzeris, “Novel
“Enhanced-Cognition” RFID architectures on organic/paper
low-cost substrates utilizing inkjet technologies,” International
Journal of Antennas and Propagation, vol. 2007, Article ID
68385, 7 pages, 2007.
[26] G. Orecchini, F. Alimenti, V. Palazzari, A. Rida, M. M.
Tentzeris, and L. Roselli, “Design and fabrication of ultra-low
cost radio frequency identification antennas and tags exploit-
ing paper substrates and inkjet printing technology,” IET
Microwaves, Antennas and Propagation, vol. 5, no. 8, pp. 993–
1001, 2011.
[27] R. Garg, P. Bhartia, L. Bahl, and A. Ittipiboon, Microstrip
Antenna Design Handbook, Artech house publishers, London,
UK, 2001.
[28] R. Ludwig and P. Bretchko, RF Circuit Design: Theory and
Applications, Prentice-Hall, 2000.
[29] CSL. CSL-461 4-Port EPC Class1 Gen2 UHF RFID Reader,
http://www.convergence.com.hk/.
[30] Tag Performance Parameters and Test Methods, EPCglobal Inc.,
2008.
doi:10.1155/2012/428284
Review Article
Some Recent Developments of Microstrip Antenna
Yong Liu, Li-Ming Si, Meng Wei, Pixian Yan, Pengfei Yang, Hongda Lu, Chao Zheng,
Yong Yuan, Jinchao Mou, Xin Lv, and Housjun Sun
Department of Electronic Engineering, School of Information and Electronics, Beijing Institute of Technology, Beijing 100081, China
Correspondence should be addressed to Yong Liu, fatufo@bit.edu.cn
Received 12 August 2011; Revised 23 November 2011; Accepted 2 January 2012
Academic Editor: Zhongxiang Q. Shen
Copyright © 2012 Yong Liu et al. This is an open access article distributed under the Creative Commons Attribution License,
Although the microstrip antenna has been extensively studied in the past few decades as one of the standard planar antennas, it still
has a huge potential for further developments. The paper suggests three areas for further research based on our previous works
on microstrip antenna elements and arrays. One is exploring the variety of microstrip antenna topologies to meet the desired
requirement such as ultrawide band (UWB), high gain, miniaturization, circular polarization, multipolarized, and so on. Another
is to apply microstrip antenna to form composite antenna which is more potent than the individual antenna. The last is growing
towards highly integration of antenna/array and feeding network or operating at relatively high frequencies, like sub-millimeter
wave or terahertz (THz) wave regime, by using the advanced machining techniques. To support our points of view, some examples
of antennas developed in our group are presented and discussed.
1. Introduction wave power increases and radiation power decreases with

the increasing substrate thickness [7], which leads to poor
The concept of microstrip antenna was first introduced in the radiation efficiency. Thus, various other techniques are pre-
1950s [1]. However, this idea had to wait nearly 20 years to
sented to provide wide-impedance bandwidths of microstrip
be realized after the development of the printed circuit board
antennas, including impedance matching networks using
(PCB) technology in the 1970s [2, 3]. Since then, microstrip
stub [9, 10] and negative capacitor/inductor [11], microstrip
antennas are considered as the most common types of
slot antennas using the U, L, T, and inverted T slots in the
antennas due to their obvious advantages of light weight,
ground plane (sometimes termed defected ground structures
low cost, low profile, planar configuration, easy of conformal,
superior portability, suitable for arrays, easy for fabrication, (DGSs)) [12, 13], surface wave suppressing using magneto-
and easy integration with microwave monolithic integrate dielectric substrate [14] and electromagnetic bandgap (EBG)
circuits (MMICs) [4–7]. They have been widely employed structures [15], and composite-resonator microstrip anten-
for the civilian and military applications such as television, nas using metamaterial resonators [16, 17]. Another problem
broadcast radio, mobile systems, global positioning system to be solved is the low gain for conventional microstrip
(GPS), radio-frequency identification (RFID), multiple- antenna element. Cavity backing has been used to eliminate
input multiple-output (MIMO) systems, vehicle collision the bidirectional radiation, thereby providing higher gain
avoidance system, satellite communications, surveillance sys- compared with conventional microstrip antenna [18]. Lens
tems, direction founding, radar systems, remote sensing, bi- covering is an alternative way to achieve gain enhancement.
ological imaging, missile guidance, and so on [8]. The lens with canonical profile, like elliptical, hemielliptical,
Despite the many advantages of typical microstrip anten- hyper-hemispherical, extended hemispherical, used to focus
nas, they also have three basic disadvantages: narrow band- the radiation beam from the radiator elements. The inte-
width, low gain, and relatively large size. The narrow band- grated lens microstrip antenna can be treated as composite
width is one of the main drawbacks of these types of antenna combined by microstrip radiator elements and
antennas. A straightforward method of improving the band- dielectric lens, which is very useful for high frequencies (mm,
width is increasing the substrate thickness. However, surface sub-mm, terahertz (THz), and optical waves) applications
[19]. It is also well known that antenna array is an effective

means for improving the gain [20–25]. Monopole
The last limitation of conventional microstrip antennas
is the relatively large size, particularly at lower microwave Reflector
frequencies, since their operation frequencies are related to
the electrical size of antenna. In general, the size of the rec- Director
tangular microstrip antenna should be of order of a half-
Ground plane
guided wavelength. This limitation was mathematically stud-
ied by Wheeler [26] and Chu [27]. There have been nu- Feeding line
merous efforts to minimize the antenna size and obtain (a) The structure of the quasi-Yagi antenna
the electrically small microstrip antenna with the raised de-
mand towards smaller and smaller wireless devices. Inductive
or capacitive loading are effective ways to reduce the size
of microstrip antennas [28]. In the former work, we dem-
onstrated that the size of microstrip antenna can be minia-
turized using composite metamaterial resonators [16, 17].
Magneto-dielectric substrates have been widely used to
miniaturize microstrip antennas due to magnetic substrates
and could provide wider bandwidths than dielectric sub-
strates [29–32]. Fractal geometries, which are composed by
self-similar structures, have opened an alternative way for
(b) The photograph of the quasi-Yagi antenna
antenna miniaturization [33].
From the above discussions, we see that many methods Figure 1: Compact broad-band quasi-Yagi antenna.
and materials are used to improve the properties of mi-
crostrip antennas. However, there should be a relationship
among bandwidth, gain, and size of the microstrip antennas.
Antenna engineers have recognized that the improvement in gain, miniaturization, circular polarization, multipolarized,
one antenna property is frequently accompanied by decline and so forth.
in its other performances. For example, the antenna size is As we know, microstrip antennas inherently have nar-
reduced usually at the expense of its bandwidth and gain. rower bandwidth and lower gain compared to conventional
Therefore, a more comprehensive consideration must be bulky antennas. Some microstrip antennas with special
given on further developments of microstrip antennas. topologies, like quasi-Yagi, planar reflector antenna, are
In this paper, we will suggest three areas for further proposed to replace the conventional bulky antennas. Here,
research based on our previous works on microstrip antenna we will take a quai-Yagi antenna as an example to show how
elements and arrays [16–25, 34–41]. We first note that novel to design a planar microstrip antenna with Yagi-Uda end-fire
microstrip antenna topologies are proposed to meet the radiation pattern. In addition, a microstrip array with special
desired requirement of variety of potential wireless appli- array topology is designed to get dual-polarized property.
cations, such as ultrawide band (UWB), high gain, minia-
turization, circular polarization, multipolarized, and so on. 2.1. Compact Broad-Band Quasi-Yagi Antenna. A novel S-
Next, we discuss the composite antennas based on microstrip band compact quasi-Yagi antenna has been designed, fabri-
antennas which have more potent than each individual cated and measured by our group, as shown in Figure 1. This
antenna. Finally, with the development of micro-/nano- antenna is composed of a printed monopole-driven element,
machining techniques, antennas/arrays with highly integra- a printed reflector element, and six printed director elements.
tion and with highly operating frequencies are discussed. We To explain the end-fire radiation behavior of the quasi-
present some examples of antennas developed in our group Yagi antenna, a comparison of radiation patterns, among
to support our points of view. (1) microstrip monopole only, (2) microstrip monopole and
a reflector, (3) microstrip monopole and a director, (4)
2. Variety of Microstrip Antenna Topologies microstrip monopole and a reflector with one director, and
(5) microstrip monopole and a reflector with six director, is
Microstrip antennas have extensively used in commercial and shown in Figure 2. We can observe that both the reflector
military applications due to their attractive advantages. How- and the director can increase the end-fire radiation, and it
ever, the traditional microstrip antennas have the impedance could be substantially improved by increasing the number of
bandwidth of only a few percent and radiation pattern with directors.
omnidirection, which obviously does not meet the require- The measured VSWR results are shown in Table 1. A
ments of various wireless applications. To this end, a wide bandwidth of 14% for VSWR less than 1.5 is achieved. The
variety of microstrip antenna topologies, including different gain of the antenna is above 7.5 dBi, as shown in Table 2. In
microstrip antenna element structures and different micros- this design, we see that the microstrip antenna with special
trip array arrangements, have been studied to meet the topology could be conveniently used to replace the bulky
desired requirement such as ultrawide band (UWB), high Yagi-Uda antenna.
Stripline feeding network
Patch
Stripline-microstrip transition
Microstrip feeding network

V
H-shape coupled slot
H
0 25 50
(mm)
(a) The structure of the dual-polarized microstrip antenna array
Figure 2: Radiation patterns of microstrip monopole only, mi-

crostrip monopole and a reflector, microstrip monopole and a
director, microstrip monopole and a reflector with one director, and (b) The photograph of the dual-polarized microstrip
microstrip monopole and a reflector with six director. antenna array
Figure 3: Dual-polarized microstrip antenna array.

Table 1: The measured VSWR of the quasi-Yagi antenna.
3
Frequency (GHz)
No.
3.25 3.5 3.75 Inband
1 1.36 1.34 1.47 <1.5
2.5
2 1.37 1.26 1.49 <1.5
3 1.36 1.25 1.48 <1.5
VSWR
2
Table 2: The measured gain of the quasi-Yagi antenna (unit: dBi).
Frequency (GHz) 1.5

No.
3.25 3.5 3.75
1 7.57 8.73 8.35
2 7.58 8.55 8.37 1
15 15.2 15.4 15.6 15.8
3 7.56 8.77 8.51 Frequency (GHz)
H-port
V-port
2.2. Dual-Polarized Microstrip Antenna Array. The dual-
polarized antenna is highly required for the radar, electronic Figure 4: The VSWR of the dual-polarized microstrip antenna
countermeasure, and aerospace systems. It is known that the array.
microstrip antenna can easily be integrated with microwave
circuits and feeding network. Here, a novel Ku-band dual-
polarization microstrip antenna array with a mixed feeding The VSWR, radiation patterns, and the isolation between
network, that is, the slot coupled feeding (V-port) and the co- two polarizations of the proposed dual-polarized microstrip
plane feeding (H-port), is designed by our group, as shown antenna array are shown in Figures 4, 5, and 6, respectively.
in Figure 3. It is a three layers structure: top microstrip patch The results indicate that this microstrip antenna array has a
layer, middle stripline feeding network layer, and bottom good impedance matching, good radiation performance, as
coplane microstrip feeding network layer. Through proper well as very high isolation (less than −25 dB), which can be
array arrangement, very good isolation can be obtained. an idea candidate for the dual-polarized wireless systems.
Figure 5: The radiation patterns of the dual-polarized microstrip antenna array at the center frequency.
Slot
Dielectric
Groove guide
Figure 7: The structure of the DCWS.
is proposed by our group to meet this requirement [36].

Figure 6: The isolation of the dual-polarized microstrip antenna
Another example of composite antenna is comprised of a
array.
dielectric lens and microstrip log-period antenna, which has
been widely applied to THz systems (this type of antenna
will be further discussed in Section 4.2). Here, we will give
3. Microstrip-Antenna-Based an example of composite antenna with “structure composite”
Composite Antenna method.
As many antenna designers have found, it is not easy to
design an antenna to meet the user-defined stringent per- 3.1. Monopulse Circular-Polarized Dielectric Complex Waveg-
formance requirements demanded by special wireless appli- uide Slot Antenna Array. Waveguide slot antenna array has
cations like military radars, surveillances, and missile guid- been widely used for wireless system, due to its advantages
ance, if only one type of antenna is considered. This dif- of high radiation efficiency, high power capacity, and high
ficulty may require the use of two more different types or reliability. However, it is hard to overcome the disadvantage
structures of antenna elements with different characteristics. of high cost of fabrication.
Composite antenna formed by two more types or structures One composite antenna with waveguide slot antenna
of antennas is particularly suitable for these applications due array property, termed dielectric complex waveguide slot
to more advantages offered by different types or structures of (DCWS), is composed with slot microstrip line and groove
antennas. For example, it is a challenging task to use single guide, as shown in Figure 7. The slot microstrip line is
type of antenna to design a dual-band dual-polarization formed by a metal clad dielectric substrate and slots etched
antenna for satellite digital multimedia broadcast (S-DMB) in the metal. This composite antenna not only maintains the
application [36]. A composite antenna composed with a advantages of the traditional waveguide slot antenna array
left-handed circularly polarized (LHCP) microstrip antenna but also has the characteristics of high consistence, easy for
and a linear polarized omnidirectional biconical antenna fabrication, and low cost.
Z
Slot array Circular polarization grid
Groove guide
Feeding network
X Y
(a) The structure of the monopulse circular-polarized DCWS antenna
array (separating view)
Figure 9: The VSWR of sum and difference port of the monopulse

circular-polarized DCWS antenna array.
(b) The photograph of the monopulse circular-polarized

DCWS antenna array.
Figure 8: Ka-band monopulse circular-polarized dielectric com-

plex waveguide slot (DCWS) antenna array.
A Ka-band monopulse circular-polarized dielectric com-

plex waveguide slot (DCWS) antenna array is designed,
fabricated, and measured by our group, as shown in Figure 8.
It consists of a circular polarization grid, a slot microstrip
array, and a groove guide and feeding network. The slot
microstrip array is fabricated on a Rogers 5880 film with
dielectric constant of 2.2 and the thickness of 0.254 mm. The Figure 10: The radiation pattern of the monopulse circular-
measured results of VSWR of sum and different port are polarized DCWS antenna array at the center frequency.
shown in Figure 9. Figure 10 shows the measured radiation
pattern at the center frequency. Some important array
performance parameters such as gain, null depth and axial
ratio (AR) are also given in Table 3. As shown in the concept was first presented in 1950s [1–3]. Clearly, the
measured results, very good performance can be obtained development of microstrip antennas is closely related with
with the DCWS antenna array. The radiating efficiency of the machining techniques. Recently, various machining tech-
the DCWS antenna array is 80%, which is almost the same niques, including multilayer printed circuit board (MPCB),
as the traditional waveguide slot antenna array. Moreover, complementary metal oxide semiconductor (CMOS), low-
the DCWS antenna array has 40% larger bandwidth than the temperature cofired ceramics (LTCC), and micro-electro-
traditional waveguide slot antenna array. mechanical systems (MEMS), are highly developed, opening
opportunities for innovative antennas, such as active anten-
4. Highly Integration and Highly Operating nas, reconfigurable antennas, metamaterial-based antennas,
Frequency Antennas Based on Advanced THz antennas, and so forth. With the availability of high-
Machining Techniques precision and high-speed advanced machining techniques,
microstrip antennas are growing towards highly integration
It is known that the microstrip antenna was first fabricated of antenna/array and feed network and operating at relatively
using PCB technology in 1970s, nearly 20 years after its high frequencies. Since they are all based on the advanced
Parasitic patch 22
Layer 1 h1 , εr1
Layer 2 h2 , εr2
Driven patch
Layer 3 h3 , εr3 21
Upper-ground coupled slot
Layer 4 h4 , εr4
Gain (dBi)
Feeding stripline
Layer 5 h5 , εr5 20
Lower ground
(a) Schematic side view of the structure of the high
integrate broadband microstrip antenna array 19
18
15.4 15.6 15.8 16 16.2 16.4 16.6 16.8 17
Frequency (GHz)
1#
2#
Figure 13: The gain of the high integrate broad-band microstrip

antenna array using MPCB technology.
Table 3: The measured data of the monopulse circular-polarized

(b) The photograph of the high integrate broadband DCWS antenna array.
microstrip antenna array
Fre. (GHz) Gain (dBi) Null depth (dB) AR (dB)
Figure 11: Ku-band high integrate broadband microstrip antenna
array using MPCB technology. f0 − 0.2 22.8 −37.3 3.8
f0 21.9 −29.9 2.9
2 f0 + 0.2 22.1 −26 4.1
1.8 4.1. High Integrate Broad-Band Microstrip Antenna Array

Using Multilayer Printed Circuit Board (MPCB) Technology.
1.6 Recently, with the development of the multilayer printed
VSWR
circuit board (MPCB) technology, the microstrip antennas

can be designed and fabricated from one-dimensional (1D)
1.4 to 2D and even 3D structures.
Based on the MPCB technology, a high integrated broad-
1.2 band Ku-band microstrip antenna array is designed, fabri-
cated, and measured by our group, as shown in Figure 11.
This antenna consists of a parasitic patch, a driven patch,
1 a stripline feeding network, a broad-band coaxial line to
15.4 15.6 15.8 16 16.2 16.4 16.6 16.8 17
Frequency (GHz) stripline transition, some buried screw holes, and some via
holes. The feeding network is integrated in the bottom of the
1# substrate of the antenna. As all of the structures fabricated at
2# once, the accuracy and the uniformity can be assured. Two
Figure 12: The VSWR of the high integrate broad-band microstrip antennas of this type are measured. The measured VSWR,
antenna array using MPCB technology. gain, and radiation pattern at the center frequency are shown
in Figures 12, 13, and 14, respectively. The measured results
show that this antenna maintains good radiation and match-
machining techniques, we suggest that a third research area ing performances with relative bandwidth of 13%. They have
of microstrip antennas is constantly introducing novel ad- also shown good uniformity by using MPCB technology.
vanced machining techniques. In the following, two exam-
ples will be presented to show how important the advanced 4.2. THz Wave Planar Integrated Active Microstrip Antenna
machining technique is to fabricate microstrip antennas. One Using Micro-Electromechanical Systems (MEMSs) Technology.
is the highly integrate broad-band microstrip antenna array THz waves typically include frequencies between 0.1 THz
fabricated using MPCB technology. Another is THz wave and 10 THz. THz technology is now becoming a promising
planar integrated active microstrip antenna using MEMS technology which has potential applications in many fields,
technology. such as short-range communication, biosensor, imaging,
−10
Gain (dB)
−20
−30
−40
−80 −60 −40 −20 0 20 40 60 80
θ (◦ )
(a) E-plane (b) H-plane
Figure 14: The radiation pattern of the high integrate broad-band microstrip antenna array using MPCB technology at the center
frequency.
integration with circuit, it is difficult to be processed in

such high-frequency regions. MEMS technology opens the
way to design of THz antennas, circuits, and systems. THz
monolithic antenna fabricated using MEMS technology and
covered by a dielectric lens, which can be considered a
composite antenna, are designed, fabricated, and measured
by our group, as shown in Figure 15.
Diodes have the functions of mixing and/or modulating
(a) The photograph of the THz monolithic antenna the carrier-wave signal. It is an effective way to reduce the
propagation path for detectors application by integrating
the diode and microstrip antenna. The extended hyper-
hemispherical dielectric lens is used to increase the gain
of the microstrip antenna. An antenna-coupled detector
integrated with a dielectric lens is designed and fabricated up
to THz range by our group. The planar microstrip log-spiral
antenna and log-period antenna have been fabricated using
micro-electromechanical systems (MEMSs) technology. The
photographs of the antennas are demonstrated in Figure 15.
The measured responses of the antenna-coupled detector
working at different frequency bands are shown in Figure 16,
(b) The photograph of the THz monolithic antenna
which can be considered to determine the effective operating
covered by a dielectric lens frequencies [19, 40]. This detector gave a valid response
from 12 GHz to 110 GHz frequencies. The results prove the
Figure 15: THz wave planar integrated active microstrip antenna validity and feasibility of the THz antenna designed using
using micro-electromechanical systems (MEMSs). micro-electromechanical systems (MEMSs ) technology.
5. Conclusion
national security, space exploration and communication,
and so forth [39–46]. To realize THz transceiver system, The advantages and disadvantages of microstrip antennas are
antenna is an essential component. We often use horn anten- discussed in this paper. In particular, three areas for further
na, lens antenna, and dielectric parabolic antenna, for THz development of microstrip antennas are presented based
systems. However, they are not easy to integrate with mono- on our previous works on microstrip antenna elements and
lithic integrate circuits. Although the microstrip antenna arrays. Variety of microstrip antenna topologies and micros-
has the merits of small volume, light weight, and easy trip-antenna-based composite antenna are discussed, and
1200
1400
Responsivity (mV/mW)
1000
1200
1000 800
800
600
600
400
12 14 16 18 20 22 26 28 30 32 34 36 38 40 42
(a) Ku-band (b) Ka-band
500 200
400
150
300
100
200
50
100
0 0
50 55 60 65 70 75 75 80 85 90 95 100 105 110 115
(c) V-band (d) W-band
Figure 16: Frequency responses test results of the THz wave planar integrated active microstrip antenna covered by a dielectric lens.
the advanced machining techniques pushing the microstrip [6] K. R. Carver and J. W. Mink, “Microstrip antenna technology,”
antennas towards the highly integration of antenna/array and IEEE Transactions on Antennas and Propagation, vol. 1, no. 1,
feeding network and the highly operating frequencies are pp. 2–24, 1981.
described. To demonstrate the distinctive features of novel [7] D. M. Pozar, “Microstrip antennas,” Proceedings of the IEEE,
microstrip antennas, various antenna elements and arrays for vol. 80, no. 1, pp. 79–91, 1992.
different applications are presented. This paper has shown [8] R. Garg, I. Bhartia, I. Bahl, and A. Ittipiboon, Micro-Strip An-
that the microstrip antennas are still very promising para- tenna Design Handbook, Artech House, Boston, Mass, USA,
2001.
digm for civilian and military wireless applications.
[9] D. M. Pozar and B. Kaufman, “Increasing the bandwidth of a
microstrip antenna by proximity coupling,” Electronics Letters,
References vol. 23, no. 8, pp. 368–369, 1987.
[1] G. A. Deschamps, “Microstrip microwave antennas,” in Pro- [10] H. F. Pues and A. R. Van de Capelle, “Impedance-matching
ceedings of the 3rd USAF Symposium on Antennas, 1953. technique for increasing the bandwidth of microstrip anten-
[2] E. V. Byron, “A new flush-mounted antenna element for nas,” IEEE Transactions on Antennas and Propagation, vol. 37,
phased array application,” in Proceedings of the Phased Array no. 11, pp. 1345–1354, 1989.
Antenna Symposium, pp. 187–192, 1970. [11] A. Kaya and E. Y. Yüksel, “Investigation of a compensated
[3] J. Q. Howel, “Microstrip antennas,” in Proceedings of the Digest rectangular microstrip antenna with negative capacitor and
of the International Symposium of the Antennas and propaga- negative inductor for bandwidth enhancement,” IEEE Trans-
tion Society, pp. 177–180, 1972. actions on Antennas and Propagation, vol. 55, no. 5, pp. 1275–
[4] I. J. Bahl and P. Bhartia, Microstrip Antennas, Artech House, 1282, 2007.
1980. [12] K. L. Wong and N. H. Hsu, “A broad-band rectangular patch
[5] J. R. James, P. S. Hall, and C. Wood, Microstrip Antenna Theory antenna with a pair of wide slits,” IEEE Transactions on Anten-
and Design, Peter Peregrinus, 1981. nas and Propagation, vol. 49, no. 9, pp. 1345–1347, 2001.
[13] S. I. Latif, L. Shafai, and S. K. Sharma, “Bandwidth enhance- [29] H. Mosallaei and K. Sarabandi, “Magneto-dielectrics in elec-
ment and size reduction of microstrip slot antennas,” IEEE tromagnetics: concept and applications,” IEEE Transactions on
Transactions on Antennas and Propagation, vol. 53, no. 3, pp. Antennas and Propagation, vol. 52, no. 6, pp. 1558–1567, 2004.
994–1003, 2005. [30] P. M. T. Ikonen, S. I. Maslovski, C. R. Simovski, and S. A.
[14] P. M. T. Ikonen, S. I. Maslovski, C. R. Simovski, and S. A. Tretyakov, “On artificial magnetodielectric loading for im-
Tretyakov, “On artificial magnetodielectric loading for improving the impedance bandwidth properties of microstrip
proving the impedance bandwidth properties of microstrip antennas,” IEEE Transactions on Antennas and Propagation,
antennas,” IEEE Transactions on Antennas and Propagation, vol. 54, no. 6, pp. 1654–1662, 2006.
vol. 54, no. 6, pp. 1654–1662, 2006. [31] P. M. T. Ikonen, K. N. Rozanov, A. V. Osipov, P. Alitalo, and
[15] R. Cocooli, F. R. Yang, T. P. Ma, and T. Itoh, “Apertu-recoupled S. A. Tretyakov, “Magnetodielectric substrates in antenna min-
patch antenna on UCPBG substrate,” IEEE Transactions on iaturization: potential and limitations,” IEEE Transactions on
Microwave Theory and Techniaues, vol. 1, no. 47, pp. 2123– Antennas and Propagation, vol. 54, no. 11, pp. 3391–3399,
2130, 1999. 2006.
[16] L.-M. Si and X. Lv, “CPW-FED multi-band omni-directional [32] P. Ikonen and S. Tretyakov, “On the advantages of magnetic
planar microstrip antenna using composite metamaterial res- materials in microstrip antenna miniaturization,” Microwave
onators for wireless communications,” Progress in Electromag- and Optical Technology Letters, vol. 50, no. 12, pp. 3131–3134,
netics Research, vol. 83, pp. 133–146, 2008. 2008.
[17] L. M. Si, H. Sun, Y. Yuan, and X. Lv, “CPW-fed com-pact pla- [33] J. Anguera, C. Puente, C. Borja, and J. Soler, “Frac-tal-
nar UWB antenna with circular disc and spiral split ring res- shaped antennas: a review,” in Wiley: Encyclopedia of RF and
onators,” in Progress in Electromagnetics Research Symposium, Microwave Engineering, K. Chang, Ed., vol. 2, pp. 1620–1635,
pp. 502–505, 2009. 2005.
[18] Y. Yuan, L.-M. Si, Y. Liu, and X. Lv, “Integrated log- [34] L. Shi, H. Sun, X. Lu, C. Deng, and Y. Liu, “Study of the
periodic antenna for Terahertz applications,” in International improved MM-Wave omni-directional microstrip antenna,”
Conference on Microwave Technology and Computational Elec- in Proceedings of the 7th International Symposium on Antennas,
tromagnetics, pp. 276–279, 2009. Propagation and EM Theory (ISAPE ’06), pp. 238–241, Octo-
[19] L. Xin, M. Jinchao, Y. Yong, Y. Weihua, N. Hongbin, and ber 2006.
G. Yafen, “Integrated active antennas in terahertz focal plane [35] L. Shi, H. Sun, and X. Lu, “A composite antenna with
imaging system,” in Proceedings of the International Workshop wide circularly polarized beamwidth,” Microwave and Optical
on Antenna Technology, pp. 157–163, 2011. Technology Letters, vol. 51, no. 10, pp. 2461–2464, 2009.
[20] Q. Xu, W. Zhang, H. Sun, and X. Lu, “Millimetre wave multi- [36] L. Shi, H. Sun, W. Dong, and X. Lv, “A dual-band multi-
polarised microstrip antenna array and application example,” function carborne hybrid antenna for satellite communication
IET Microwaves, Antennas and Propagation, vol. 4, no. 10, pp. relay system,” Progress in Electromagnetics Research, vol. 95, pp.
1525–1530, 2010. 329–340, 2009.
[21] M. Wei, H. Deng, H. Sun, and Y. Liu, “Design of an X/Ka [37] Y. Chen, H. Sun, and X. Lv, “Novel design of dual-polarization
dual-band co-aperture broadband microstrip antenna array,” broad-band printed L-shaped probe fed microstrip patch
in Proceedings IEEE International Conference on Microwave antenna for SAR applications,” in IET International Radar
Technology and Computational Electromagnetics, pp. 217–220, Conference 2009, pp. 1–4, April 2009.
2011. [38] L. M. Si, H. J. Sun, and X. Lv, “Numerical simulations
[22] Y. Liu, X. Lu, Y. Yuan, Y. Chen, and L. Shi, “Integrated design of backward-to-forward leaky-wave antenna with composite
and research of Ka-band electronically large mono-pulse right/left-handed coplanar waveguide,” Chinese Physics Letters,
planar antenna and feed system,” in Proceedings of the 7th vol. 27, no. 3, Article ID 034106, 2010.
International Symposium on Antennas, Propagation and EM [39] L.-M. Si, H.-J. Sun, and X. Lv, “THz leaky-wave antenna with
Theory (ISAPE ’06), pp. 150–153, October 2006. high-directivity and beam-steering using CPW CRLH meta-
[23] Y. Liu, H. J. Sun, X. Lu et al., “Study of the millimeter material resonators,” in Proceedings of the 3rd International
microstrip monopulse antenna array in the dual-sensor Symposium on Photoelectronic Detection and Imaging, Beijing,
system,” in Proceedings of the 7th International Symposium on China, 2009.
Antennas, Propagation and EM Theory (ISAPE ’06), pp. 157– [40] Y. Liu, L. M. Si, S. Zhu, and H. Xin, “Experimental realization
160, October 2006. of an integrated THz electromagnetic crys-tals (EMXT) H-
[24] Y. An, X. Lv, and B. Gao, “Developing a kind of mi-crostrip plane horn antenna,” Electronics Letters, vol. 47, no. 2, pp. 80–
array antenna with beam squint,” in Proceedings of the 5th 82, 2011.
International Symposium on Antennas, Propagation and EM [41] L. M. Si, Y. Liu, S. H. Zhu, and H. Xin, “Integrated THz
Theory, pp. 443–446, 2000. horn antenna using EBG structures,” in Proceedings of the IEEE
[25] L. Yong, L. Xin, H. Zhao, and Y. Yuan, “Integrated design and International Symposium on Antennas and Propagation and
research of novel KA-band circular polarized monopulse inter- USNC/CNC/URSI National Radio Science Meeting, University
rogator array antenna,” in Proceedings of the IET International of Colorado at Boulder, 2011.
Radar Conference 2009, pp. 1–4, April 2009. [42] Y. Liu, Y. Yuan, X. Lv, and L.-M. Si, “Design of 0.5 THz
[26] H. A. Wheeler, “Fundamental limitations of small an-tennas,” 2D square lattice EBG waveguide transmission line and
Proceedings of the IRE, vol. 35, pp. 1479–1484, 1947. power-divider using MEMS technology,” in Proceedings 3rd
[27] L. J. Chu, “Physical limitations of omni-directional antennas,” International Symposium on Photoelectronic Detection and
Journal of Applied Physics, vol. 19, no. 12, pp. 1163–1175, 1948. Imaging, vol. 7385, Beijing, China, 2009.
[28] R. Azadegan and K. Sarabandi, “A novel approach for minia- [43] L.-M. Si, Y. Yong, H.-J. Sun, and L. Xin, “Characterization and
turization of slot antennas,” IEEE Transactions on Antennas application of planar terahertz narrow bandpass filter with
and Propagation, vol. 51, no. 3, pp. 421–429, 2003. metamaterial resonators,” in Proceedings of the International
Workshop on Metamaterials, pp. 351–354, Nanjing, China,

2008.
[44] L. M. Si and X. Lv, “Terahertz waves hairpin microstrip band-
pass filter and its application to overlaid dielectric material
detection,” Modern Physics Letters B, vol. 22, no. 29, pp. 2843–
2848, 2008.
[45] L.-M. Si, H.-J. Sun, and X. Lv, “Theoretical investigation of
terahertz amplifier by carbon nanotubes within transmission
line metamaterials,” Microwave and Optical Technology Letters,
vol. 53, no. 4, pp. 815–818, 2011.
[46] W. Hong, “Millimeter wave and THz communications in
China,” in Proceedings of the IEEE International Microwave
Symposium Digest (MTT ’11), pp. 1–4, 2011.
doi:10.1155/2012/786791
Research Article
New Configurations of Low-Cost Dual-Polarized Printed
Antennas for UWB Arrays
Guido Valerio,1 Simona Mazzocchi,2 Alessandro Galli,2

Matteo Ciattaglia,3 and Marco Zucca3
1 Institut
d’Electronique et des Télécommunications de Rennes (IETR), UMR CNRS 6164, Université de Rennes 1,
35042 Rennes Cedex, France
2 Department of Information Engineering, Electronics and Telecommunications, Sapienza University of Rome, 00184 Rome, Italy
3 SELEX S.I. S.p.A., Sistemi Radianti, 00131 Rome, Italy
Correspondence should be addressed to Guido Valerio, guido.valerio@univ-rennes1.fr
Received 15 August 2011; Revised 26 November 2011; Accepted 6 December 2011
Academic Editor: Athanasios Panagopoulos
Copyright © 2012 Guido Valerio et al. This is an open access article distributed under the Creative Commons Attribution License,
A novel class of structures is proposed to realize ultra-wide-band radiating elements for large arrays, providing dual polarization,
beam scanning, and compact and inexpensive realization based on suitable rhombic arrangements of dipoles printed on low-cost
layered substrates. In a first implementation, four rhombic shapes, orthogonally placed on the same layer, provide two orthogonal
polarizations. In a second implementation, the two polarizations are excited by two rhombic shapes printed on two different layers
in a stacked-patch-like arrangement. This latter structure leads to a better lateral shielding of the single radiating element, in order
to reduce mutual interactions among adjacent elements in array environment. The behavioral features of these antennas have
been tested with various parametric analyses. Practical aspects have been addressed such as the choice of appropriate feeding and
of commercially available dielectric layers. The resulting antennas are matched at the input ports in an extremely wide range of
frequencies (5–25 GHz), covering various microwave applications, such as aircraft surveillance, weather polarimetric radars, and
control and communications systems. Good radiating features, in terms of pattern shape and gain, are observed in a large band of
frequencies. The basic scanning performance of large and small array configurations is finally investigated.
1. Introduction Many UWB antennas based on printed configurations

have been proposed in the past years [3, 10]. They are cur-
The development of ultra-wide-band (UWB) antennas is rently of great interest due to various advantages [11, 12]: low
recently providing new solutions for the design of innovative, cost, light weight, compactness and conformability, compat-
versatile, and economical radar and communication systems. ibility with printed-circuit-board (PCB) technologies, and
This class of structures is in fact gaining increasing attractive- high reproducibility in the realization of large array radars.
ness for the enhanced features of the wideband signals, for
The most common implementation is based on monopole
instance in applications to imaging, surveillance, and high-
structures, having bidirectional radiating properties. While
capacity communications [1–4].
In impulse radar techniques, UWB antennas allow to a reflecting ground plane could be introduced to maximize
transmit and receive short-pulsed signal waveforms without the gain in the required half plane, this usually introduces
significant distortions. In the field of wireless communica- resonant behaviors in the currents on the metalization, thus
tions, the demand for wideband radiating elements grows significantly reducing the useful operational bandwidth of
up as well. Furthermore, the dramatic reduction of costs the structure.
related to a single equipment for multiple applications More complex configurations should be designed if
(communications, surveillance, weather, etc.) is possible if a grounded structures are of interest: for example, the use
suitably wide band of frequency can be used and efficiently of parasitic elements can induce resonances at differ-
treated by the same radiating element [5–9]. ent frequencies thus enlarging the total available band
[13–15]. In this paper, novel configurations of printed

dipoles arranged in rhombic shapes are proposed to radiate
over a very wide band of frequency with applications to
array radars and UWB communication systems, following
a preliminary design presented in [16]. The frequency
ranges of interest here are mainly for the C, X, Ku , and
K bands (approximately between 5 and 25 GHz). Suitable
scaling of such structures anyway allows for covering other z
y
possible microwave ranges for UWB applications. The basic
principles for the new types of radiators are introduced in x
Section 2. Specific solutions are considered here for dual- (a)
polarization applications. In Section 3, a first topology is
Port
described, composed by four coplanar rhombic elements 3
(see Figure 1). In Section 4, a laterally shielded structure, 3
feasible for implementation in an array environment, is 2
1 4
presented: two rhombic elements are printed on different 5
layers, in a simple stacked configuration, each providing a
linear polarization orthogonal to each other (see Figure 2).
Extensive results are shown for both the structures, to show
the UWB matching features through the relevant input
parameters, and the radiating characteristics of the single Port Port
2 1
radiating elements and of various array configurations.
2. Basic Structures
The two structures presented here are based on various y
arrangements of printed dipoles of different lengths forming
approximately rhombic-shaped elements (see Figures 1 and
x
2). The central patch is the longest and can be fed at its ends Port
4 z
with one or two probes; the side patches, being six or eight
depending on the structure, have smaller dimensions and are (b)
parasitic. To further increase the bandwidth, the dipoles are
Figure 1: Geometry of the radiating element under analysis in
printed on a three-layer structure, designed with low-cost
Section 3, based on four rhombic coplanar elements printed on
commercially available dielectric substrates. In the present
the top of a three-layer configuration. (a) 3D view of the antenna.
analysis, the lower and the upper layers are duroid RT 5880 (b) Upper view of the structure. Lowest substrate: thickness h =
Rogers substrates with relative permittivity εr,1 = 2.2 and 1.5 mm, relative dielectric constant εr = 2.2, loss tangent tan δ =
dielectric loss tangent tan δ1 = 9.0 · 10−4 , the middle layer 9 · 10−4 (duroid RT 5880 Rogers). Middle substrate thickness h =
is an FR4 epoxy substrate with εr,2 = 4.4 and dielectric 7.5 mm, relative dielectric constant εr = 4.4, loss tangent tan δ =
loss tangent tan δ2 = 200.0 · 10−4 . It should be pointed 200 × 10−4 (FR4 epoxy). Uppermost substrate thickness h = 2 mm,
out that, even if loss effects can become nonnegligible relative dielectric constant εr = 2.2, loss tangent tan δ = 9 · 10−4 (RT
(particularly in higher-frequency ranges); in this case the 5880 Roger). Overall unit-cell dimension 43 × 43 mm. Geometrical
choice of such inexpensive commercial substrates is mainly parameters of the patches: w1 = 1.9 mm, l1 = 17.3 mm, w2 = 2 mm,
related to the strong reduction of costs in the implementation l2 = 7 mm, w3 = 1 mm, l3 = 11.3 mm, w4 = 0.6 mm, l4 = 4 mm,
of large arrays of elements, exploiting also well-established w5 = 0.4 mm, and l5 = 2 mm (wi and li are the width and the length
of the ith patch, resp.). Diameter of the probe d = 0.86 mm.
manufacturing printed-circuit techniques of PCB [17]. A
number of other materials having similar electric parameters
but reduced losses can be chosen if higher efficiency is desired
(e.g., for the internal layer a dielectric such as TMM4, having a couple of rhombic elements, mutually rotated of 90◦ . The
a loss tangent one order of magnitude lower than FR4 can be relevant features of two possible rhombic arrangements are
employed). Some compared results concerning the influence analyzed in detail in the next two sections.
of such losses on the gain and efficiency of these antennas will
be presented next. In the following analyses, also nonideal
effects of the metalization are taken into account, considering 3. Unshielded Antenna with Coplanar
finite-conductivity strips made of copper (σ = 5.8 · 107 S/m, Dual Rhombic Elements
μr = 0.999991) having nonzero thickness (10−6 m). A single
rhombic shape mainly provides a field linearly polarized, 3.1. Structure. A fully dual-polarized radiating element can
with components related to the direction of the relevant be obtained by arranging four rhombic shapes, printed on
dipoles. The dual polarizations can thus be obtained using the top of the three-layer structure described in the previous
operational requirements are given at lower and at higher

frequencies. Specifically, at lower frequencies both a good
matching and a regular radiation pattern are required, whilst
at higher frequencies, because of a different utilization of
the control system, a good matching only can be sufficient.
The dimensions of the various patches have been designed in
order to reach an optimum input matching in the working
z wide band 5–25 GHz, as shown in next subsection. Further
y simulations, here omitted for brevity, would show that the
input matching is preserved also at frequencies higher than
x
30 GHz, even though the radiation patterns are less regular
(a) as the frequency increases.
Port Beyond these advantages in terms of extremely wide
3 band, low-profile and low-cost realization, and radiation
4
3
pattern regularity, a key promising feature of this structure
1
2 2 3 4 is also the dynamic conformability of its radiation pattern,
Port 1 Port due to the use of a pair of probes for the feeding of each
2 1 polarization. For instance, if an array of elements is placed on
a conformable surface, a fine tuning of the pointing angle can
be easily obtained by an appropriate tailoring of the feeding
y Port
network.
4
x
3.2. Analysis and Results. A throughout optimization pro-
z
cedure, involving the dimensions of the patches and the
(b) positions of the feeding probes, has been carried out with
the software package “ModeFrontier” [18], with the aim of
Figure 2: Geometry of the radiating element under analysis in
Section 4, based on another rhombic element printed on the top achieving the best input matching in the desired frequency
layer, and a rhombic element printed on the second layer. (a) 3D band. The optimization has been performed at first on a
view of the antenna. (b) Upper view of the structure. Same three- single rhombic shape and then refined with the full four-
layer configuration as in Figure 1. Geometrical parameters of the rhombus structure. The algorithm used is a multiobjective
patches of the upper element: w1 = 1 mm, l1 = 29.6 mm, w2 = genetic algorithm with multisearch elitism for enhanced
0.8 mm, l2 = 10 mm, w3 = 0.6 mm, l3 = 8 mm, w4 = 0.6 mm, and robustness [19]. Its objectives were chosen as the conditions
l4 = 7 mm. The diameter of the probes feeding the upper element |Si j | < −10 dB at 5 GHz for any i, j, and the variables were all
is d = 0.24 mm. Geometrical parameters of the patches of the lower the dimensions and positions of the various patches.
element: w1 = 0.6 mm, l1 = 26.8 mm, w2 = 0.6 mm, l2 = 15.6 mm, The results shown here have been computed with the
w3 = 0.6 mm, l3 = 6 mm, w4 = 0.6 mm, and l4 = 6 mm. (wi and li time-domain solver of the high-frequency electromagnetic
are the width and the length of the ith patch, resp.) The diameter of
CAD “CST Studio 2010” [20]. The layered substrate is
the probes feeding the lower element is d = 0.4 mm. The size of the
box is 40 mm.
assumed laterally unbounded, and a single radiating element,
composed by four rhombic shapes, is considered as in
Figure 1. The input ports are modeled as coaxial cables,
fed with proper phase shifts. The feeding network providing
section, as in Figure 1. The central patch of each rhombic these shifts among the different ports is not simulated
element is fed with a probe at its outer end, and opposite here and will be object of future work. The CST model is
patches are fed with signals having the same magnitude discretized with hexahedral cells of average dimension λ/10;
and opposite sign. With this geometrical arrangement, one open conditions are placed at the side boundary of the cell,
couple of opposite rhombic elements provides one linear thus assuming a laterally unbounded substrate; extra space is
polarization, and the other couple of opposite elements, added in the top half plane in order to accurately estimate the
rotated of 90◦ , provides the orthogonal polarization. If the radiation patterns. Waveguide ports are defined at the coaxial
same signal is radiated through both the polarizations, a cables and excited with the fundamental TEM mode.
general elliptical polarization can be obtained, by suitably Based on the data sheets of the electromagnetic param-
tuning the relevant phase shifts among the input probes. eters of these materials as a function of the frequency (if
In particular, a circular polarization can be obtained by available), the dispersive effects can be taken properly into
imposing a 90◦ phase shift between adjacent elements. account in the simulations. The relevant analysis shows
The antenna is designed to work in a very wide frequency anyway that these dispersive effects do not change sensitively
range, possibly greater than C-X bands (4–12 GHz) in the the performance of our antennas in the ranges of interest.
case of interest. Its transverse dimensions of the element Due to the geometrical symmetries, the input features
are approximately 4 cm, leading to a possible compact of the antenna are described by three different scattering
implementation in array at these frequencies. Different parameters. In Figure 3, the magnitude in dB of these three
0 0
10
30 30
−10 0
−20 −10
(dBi)
60
−20
−30
−30
−40
−40 90
−50
(a)
−60 0
10
0 5 10 15 20 25 30 30
f (GHz) 0
| S11 | (dB) −10
(dBi)
| S21 | (dB) 60
| S31 | (dB) −20
−30
Figure 3: Antenna of Figure 1. Magnitude of the scattering
coefficients of the radiating element fed by the four probes.
−40 90
Reflection coefficient |S11 | at the input port 1 (black line) and
transmission coefficients |S21 | from port 1 to port 2 (gray line) and (b)
from port 1 to port 3 (dashed black line).
Figure 4: Antenna in Figure 1. Gain in polar form on the principal
elevation plane φ = 0◦ of the single element at different frequencies
in the operational bandwidth 5–25 GHz: 5 GHz (solid black line),
parameters is shown in the frequency band considered, from 10 GHz (solid gray line), 20 GHz (dashed black line). (a) Ports 1 and
dc to 25 GHz. The reflection coefficient at one input cable 2 are fed, exciting a field along the θ direction. (b) Ports 3 and 4 are
fed, exciting a field along the ϕ direction.
is shown in black solid line (due to the symmetry of the
structure, the cable considered is arbitrary, since S11 = S22 =
S33 = S44 ). The coupling between opposite ports (e.g.,
|S21 |) is shown in gray solid line, and the coupling between 4. Shielded Antenna with Stacked
adjacent ports (e.g., |S31 |) is shown in black dashed line. The Dual Rhombic Elements
operational frequency band where the parameters are under
the conventional threshold of −10 dB is extremely wide: in 4.1. Structure. Once the antenna described in the previous
the figure, the band 5–25 GHz is well matched, but further section is introduced in an array environment, the UWB
simulations would show a good matching also beyond input performance of the array could deteriorate due
30 GHz, even though, as said, the radiation patterns tend to to strong mutual coupling among adjacent elements. To
deteriorate as the frequency increases. In Figure 4(a), the gain overcome such a problem, a lateral shielding of the single
is shown on the principal plane ϕ = 0◦ at different frequen- radiating element can be implemented with metallic walls.
cies from 5 GHz to 25 GHz if the ports 1 and 2 are fed with Such a configuration also enables us to reduce adverse effects
common amplitude and opposite phase: by symmetry, the related to possible launching of surface and leaky waves in the
electric field has only a θ-component. In Figure 4(b), the gain layered structure [11, 12]. As an example of this approach,
is shown on the same plane if the ports 3 and 4 are fed with an alternative shielded radiating element is presented here,
common amplitude and opposite phase: in this case, by sym- based on the same rhombic printed shapes described above,
metry, the electric field has only a ϕ-component. Good radi- but with elements placed on different layer interfaces.
ation features are found in the lower band of frequency, while In this case, shown in Figure 2, each couple of opposite
the pattern tends to deform at higher frequencies as expected. rhombic elements is here replaced by a single rhombus,
The features of the maximum gain are resumed in whose central larger patch is fed by two different probes at its
Table 1, considering their values with respect to the isotropic ends, carrying signals with common amplitude and opposite
radiator (dBi and the relevant beam angular locations on phase. With this original feeding structure involving two
the different planes). In the same table, maximum gain and probe elements on the opposite sides of each main strip, it is
radiation efficiencies are also presented with reference to noted that, by properly choosing the phase shift between the
the alternative choice of the TMM4 for the middle layer feeders, it is also possible to achieve a straightforward control
(with reduced losses, see Section 2 for details). As shown, of the current configuration excited on the strips, thus
acceptable results are achieved even with the lossy substrate, suitably influencing the directional features of the radiated
while significant improvements in the efficiency can be beam. In this structure, the two orthogonal polarizations are
obtained at higher frequencies with the use of a lower-loss then provided by two different rhombic elements, rotated
substrate. by 90◦ and printed on two different layers: the dipoles of
Table 1: Maximum gain: on ϕ = 0◦ .
f (GHz) Direction (ϑ) Max gain Efficiency Max gain (low loss) Efficiency (low loss)
5 ϑ = 0◦ 6.6 dBi 92.9% 7.1 dBi 99%
10 ϑ = 90◦ , ϕ = 54◦ 3.8 dBi 78.7% 4.5 dBi 96.6%
15 ϑ = 45◦ , ϕ = 22◦ 3.2 dBi 80.0% 3.7 dBi 96.8%
20 ϑ = 38◦ , ϕ = 24◦ 3.7 dBi 64.1% 4.9 dBi 92.7%
25 ϑ = 90◦ , ϕ = 9◦ 5.5 dBi 68.4% 6.2 dBi 93.3%
one rhombus are printed on the uppermost layer and the 0

dipoles of the other rhombus on the middle layer. Also in
−10
this case the box has transverse dimensions of 4 cm, leading
to good performance in a possible array implementation. On
−20
the other hand, the distance between the structure (i.e., strips
and cables) and the lateral boundaries is greater in this new
(dB)
−30
configuration than in the previous one, leading to an easier
input matching also in the shielded configuration. −40
Again, the dimensions of the various patches have been
optimized in order to yield the optimum input matching in −50
the frequency band 5–25 GHz. Of course, considering such a
different arrangement of the rhombic shapes, rather different −60
geometrical parameters have been reached in this second 0 5 10 15 20 25
structure as optimum values. f (GHz)
| S11 |
4.2. Analysis and Results. The results shown here have | S33 |
been obtained through full-wave simulations of the final | S13 |
optimized stacked and shielded structure again with the
time-domain solver of CST Studio. Figure 5: Antenna of Figure 2. Magnitude of the scattering
In Figure 5, the magnitude in dB of the three scattering coefficients of the radiating element fed by the four probes.
parameters of the shielded configuration is shown; also with Reflection coefficients |S11 | at the input port 1 feeding the upper
structure (black line), |S33 | at the input port 3 feeding the lower
this structure, a good input matching is reached in the very
structure (gray line), and transmission coefficient |S31 | from port
wide frequency band 5–25 GHz. The reflection coefficient 1 to port 3 (dashed black line).
at port 1 (i.e., feeding the upper rhombic shape) is shown
in solid black line, the reflection coefficient at port 3 (i.e.,
feeding the lower rhombic shape) is shown in solid gray line,
while the coupling between the rhombic shapes is shown in In Tables 2 and 3, a summary of the maximum-gain
dashed black line. values, locations, and radiation efficiencies is also given
The radiation patterns on the principal planes are shown for different frequencies, with reference to the feeding
again for various values of the frequency. In Figure 6(a), of the upper and lower element, respectively. As for the
the gain is shown on the principal plane ϕ = 0◦ , when previous unshielded antenna, maximum gain and efficiency
probes 1 and 2 feed the upper element with signals with is also presented with reference to the alternative choice of
common magnitude and 180◦ phase shift. The electric the TMM4 for the middle layer (with reduced losses, see
field is in this case polarized along the θ direction by Section 2 for details). In particular, while the lower element
symmetry. In Figure 6(b), the gain is shown on the same has a reduced efficiency at higher frequencies, its behavior
plane when the lower element is fed through probes 3 and can be substantially improved with the use of a lower-loss
4, with common magnitude and 180◦ phase shift. A dual substrate.
polarization is radiated with respect to the previous result,
the electric field being polarized along the ϕ direction. As 4.3. Array Behavior. In order to test further this type
seen, fairly regular radiation patterns are observed for both of structure, a first simple analysis has been led which
polarizations, in particular in the lower part of frequency. gives basic information on the scanning-beam directional
Since this configuration is studied with an “open add space” features for large arrays. To this aim, the radiation pattern
lateral boundary in CST, an estimation of backlobes is also of an array of 140 × 140 elements has been computed
present in the results. with an array-factor approximation, for different values of
The gain on the other principal plane ϕ = 90◦ is shown the pointing angle, depending on the selected phase shift
in Figure 7. The upper rhombic shape is fed in Figure 7(a), between adjacent elements of the array. In Figure 8(a), the
while the lower rhombic shape is fed in Figure 7(b). pattern on the plane ϕ = 0◦ at 5 GHz is shown for a
20 0 0
30 20
30 30 30
10 10
0 0
(dBi)
60
(dBi)
−10 −10 60
−20 −20
−30 −30
−40 90 −40 90
120 120 120 120
150 150 150 150

180 180
(a) (b)
Figure 6: Antenna in Figure 2. Gain in polar form on the principal elevation plane φ = 0◦ of the single element at different frequencies
in the operational bandwidth 5–25 GHz: 5 GHz (solid black line), 10 GHz (gray line), 20 GHz (dashed black line). (a) Ports 1 and 2 feed the
upper element, exciting a field along the θ direction. (b) Ports 3 and 4 feed the lower element, exciting a field along the ϕ direction.
0 0
20 20
30 30 30 30
10 10
0 0
(dBi)
(dBi)
−10 60 −10 60
−20 −20
−30 −30
−40 90 −40 90
120 120 120 120
150 150 150 150

180 180
(a) (b)
Figure 7: Antenna in Figure 2. Gain in polar form on the principal elevation plane φ = 90◦ of the single element at different frequencies
in the operational bandwidth 5–25 GHz: 5 GHz (solid black line), 10 GHz (gray line), 20 GHz (dashed black line). (a) Ports 1 and 2 feed the
upper element, exciting a field along the θ direction. (b) Ports 3 and 4 feed the lower element, exciting a field along the ϕ direction.
broadside radiation when the four probes of each element in connection of the element spacing as phase shift and
are fed with common amplitude and a 90◦ shift in order frequency are varied.
to radiate a circular polarized field. The two components Further results have been obtained considering a linear
along θ and ϕ are shown in solid black line and in dashed array made of a small number of cells (3 × 1) along the x
gray line, respectively. In Figure 8(b), the same quantities direction in the reference system of Figure 2. In this case, to
are computed for a beam pointing at the elevation θ = 30◦ accurately predict the array performance, a nonapproximate
and azimuth ϕ = 0◦ . The gain at the main lobe direction is full-wave analysis has been necessary with a proper CAD
fairly regular, around 50 dBi; the side-lobe levels are rather implementation of the overall physical structure of the three
reduced (about 20 dB below the main lobe); the considered antennas. This rigorous approach allows us to to verify
phase-scanned pencil beams have half-power beamwidth of the actual scanning features of the radiated beam as a
about 0.4◦ . Effects related to grating lobes can be present function of the relevant phase shift. Representative behaviors
Table 2: Maximum gain if ports 1 and 2 are fed (Upper Element).
5 ϑ = 0◦ 6.5 dBi 92.9% 6.9 dBi 99.3%
10 ϑ = 0◦ 6.5 dBi 56.1% 10.67 dBi 96.8%
15 ϑ = 55◦ , ϕ = 70◦ 4.2 dBi 47.3% 7.01 dBi 88.5%
20 ϑ = 50◦ , ϕ = 67◦ 2.5 dBi 40.6% 9.04 dBi 75.0%
25 ϑ = 21◦ , ϕ = 90◦ 4.2 dBi 41.5% 7.63 dBi 86.9%
Table 3: Maximum gain if ports 3 and 4 are fed (lower element).
5 ϑ = 0◦ 7.5 dBi 79.0% 8.4 dBi 95.8%
10 ϑ = 0◦ 8.2 dBi 37.0% 10.3 dBi 88.1%
15 ϑ = 61◦ , ϑ = 20◦ 2.5 dBi 28.0% 6.8 dBi 80.0%
20 ϑ = 0◦ 2.3 dBi 28.2% 7.8 dBi 88.6%
25 ϑ = 45◦ , ϑ = 20◦ 3.4 dBi 29.1% 7.5 dBi 80.8%
60 60
40 40
20 20
(dBi)
(dBi)
0 0
−20 −20
−40 −40
−90 −45 0 45 90 −90 −45 0 45 90
θ (◦ ) θ (◦ )
Abs (θ) [dBi] Abs (θ) [dBi]

Abs (ϕ) [dBi] Abs (ϕ) [dBi]
(a) (b)
Figure 8: Array of shielded elements as in Figure 2. Gain on the principal elevation plane φ = 0◦ of a 140 × 140 array at 5 GHz. All the
ports are fed, exciting a circularly polarized field (its θ component in solid black line, its ϕ component in dashed gray line). (a) Beam pointing
at broadside (θ = 0◦ ). (b) Beam pointing at θ = 30◦ .
of radiation patterns for this small array are reported in the other cables are shown, avoiding coefficients having equal
Figure 9. In Figure 9(a), the gain on the two principal planes values due to evident geometrical symmetries. As expected,
is shown when the three upper sets of patches are fed in phase in the band of frequency investigated, a low level of coupling
(black curves) and when three lower sets of patches are fed in is found also in this 3 × 1 array. In Figure 10(b), the three
phase (gray curves). In Figure 9(b), the scanning capabilities lower sets of patches of the array are fed in phase through
of this small array are presented, showing the gain on the one of the two ports; the magnitude of the active reflection
principal plane ϕ = when the lower sets of patches are fed, coefficient at the middle cell and at one side cell is reported,
pointing their main beam along the ϕ = 0◦ plane, at θ = 10◦ , proving a good input matching also in this active array
20◦ , 30◦ , 40◦ . As expected, a fan beam is obtained, as typical configuration.
of linear arrays, having maximum gain between 12 and 9 dBi. As a last result, in Figure 11 the radiated patterns of
In Figure 10, couplings among cells are analyzed, again an array made of 3 × 3 elements is shown on both the
with reference to the feeding of the lower patches, only for the principal planes, ϕ = 0◦ , 90◦ for both the upper (Figure 11(a))
sake of brevity. In Figure 10(a), an input cable feeds the lower and lower (Figure 11(b)) patches. Again, a full-wave analysis
patch only in the central cell. The magnitude of the reflection has been performed, without recurring to simplified array-
coefficient at this port and of the coupling coefficients with factor formulations. All the cells of the array are fed in phase
20 60
10 40
0 20
(dBi)
(dBi)
−10 0
−20 −20
−30 −40
−90 −60 −30 0 30 60 90 −90 −60 −30 0 30 60 90
θ (◦ ) θ (◦ )
(a) (b)
Figure 9: Array of 3 adjacent shielded cells as in Figure 2 at 5 GHz. (a) Gain on the principal elevation planes φ = 0◦ (solid lines) and φ = 90◦
(dashed lines). Beam pointing at broadside (θ = 0◦ , all the cells are excited in phase) by the upper patches (black lines) and by the lower
patches (gray lines). (b) The three lower sets of patches are phased in order to point at θ = 10◦ (black line), θ = 20◦ (gray line), θ = 30◦ (black
line with squares), and θ = 40◦ (gray line with squares).
0 0
−5
−10
−10
−20
−15
(dB)
(dB)
−30 −20
−25
−40
−30
−50
−35
−60 −40
5 10 15 20 25 5 10 15 20 25
f (GHz) f (GHz)
(a) (b)
Figure 10: Array of 3 adjacent shielded cells as in Figure 2 at 5 GHz. (a) Only the input port 3 feeds the lower patches in the central cell.
Magnitude of reflection coefficient (thick black line), of the coupling coefficient with a cable connected to the upper patch in the same cell
(thick gray line), and of coupling coefficients with cables in an adjacent cell (thin black lines). (b) All the three lower sets of patches as fed in
phase through port 3 (see Figure 2) and the respective translated ports in the other two cells. Magnitude of the active reflection coefficient
versus frequencies at the middle cell of the 3 × 1 array (black line) and at one side cell (gray line).
thus radiating a single beam at broadside, with gain around orientation. Attention was paid to practical implementa-
17 dBi, angular width 20◦ , and sidelobe level −15 dB with tion with inexpensive dielectrics commonly used in PCB
respect to the maximum gain. As shown in the pictures, a technology. These structures are low profile and lightweight
remarkable symmetry of the pattern is obtained on the prin- and are characterized by high modularity/scalability, which
cipal planes, also if the different polarizations are compared. makes them suitable to implement low-profile phased array
antennas of various shapes and sizes.
5. Conclusion Suitable extensive parametric analyses have been carried
out by means of advanced numerical tools as concerns the
A new class of UWB low-cost printed antennas has been pre- most efficient choice of the strip geometry configurations.
sented and optimized for dual-polarized radar applications A first design of the antenna element is made with four
in the microwave range 5–25 GHz. The basic single element coplanar rhombic elements, each fed by a probe reaching one
is composed by properly fed printed dipoles arranged in end of the central dipole. Relevant results are shown for a
rhombic configurations. Different elements can provide single element printed on a laterally unbounded substrate.
almost orthogonal polarizations, depending on their mutual An advanced alternative design has been proposed, leading
20 [3] C.-C. Chen, “Ultrawide bandwidth antenna design,” in

Antenna Engineering Handbook, J. L. Volakis, Ed., McGraw-
Hill, New York, NY, USA, 2007.
10
[4] J. J. Lee, “Ultra wideband arrays,” in Antenna Engineering
Handbook, J. L. Volakis, Ed., McGraw-Hill, New York, NY,
0 USA, 2007.
(dBi)
[5] J. Herd, S. Duffy, M. Weber, G. Brigham, C. Weigand, and D.

−10 Cursio, “Advanced architecture for a low cost Multifunction
Phased Array Radar,” in Proceedings of the IEEE MTT-S
International Microwave Symposium Digest (MTT ’10), pp.
−20
676–679, Anaheim, Calif, USA, May 2010.
[6] M. Weber, J. Cho, J. Flavin, J. Herd, and M. Vai, “Multi-
−30 function phased array radar for U.S. Civil-sector surveillance
−90 −60 −30 0 30 60 90 needs,” in Proceedings of the 32nd Conference on Radar
θ (◦ ) Meteorology, pp. 1977–1987, Albuquerque, NM, USA, October
(a) 2005.
[7] M. Weber, J. Cho, J. Flavin, and J. Herd, “Multifunction
20
phased array radar: technical synopsis, cost implications and
operational capabilities,” in Proceedings of the 23rd Conference
10 on International Interactive Information and Processing Systems
(IIPS) for Meteorology, Oceanography and Hydrology, San
Antonio, Tex, USA, January 2007.
0
[8] R. Waterhouse, Ed., Printed Antennas for Wireless Communi-
(dBi)
cations, Wiley, New York, NY, USA, 2007.

−10
[9] V. N. Bringi and V. Chandrasekar, Polarimetric Doppler
Weather Radar: Principles and Applications, Cambridge Uni-
−20 versity Press, Cambridge, UK, 2001.
[10] D. R. Jackson, “Microstrip antenna,” in Antenna Engineering
Handbook, J. L. Volakis, Ed., McGraw-Hill, New York, NY,
−30
−90 −60 −30 0 30 60 90 USA, 2007.
θ (◦ ) [11] C. D. Nallo, F. Mesa, and D. R. Jackson, “Excitation of leaky
modes on multilayer stripline structures,” IEEE Transactions
(b) on Microwave Theory and Techniques, vol. 46, no. 8, pp. 1062–
1071, 1998.
Figure 11: Array of 3 × 3 adjacent shielded cells as in Figure 2. Gain
on the principal elevation planes φ = 0◦ (solid lines) and φ = 90◦ [12] P. Baccarelli, P. Burghignoli, F. Frezza, A. Galli, G. Lovat,
(dashed lines) at 5 GHz. Beam pointing at broadside (θ = 0◦ , all the and D. R. Jackson, “Approximate analytical evaluation of
cells are excited in phase). (a) The upper patches are fed. (b) The the continuous spectrum in a substrate-superstrate dielectric
lower patches are fed. waveguide,” IEEE Transactions on Microwave Theory and
Techniques, vol. 50, no. 12, pp. 2690–2701, 2002.
[13] K. Ghorbani and R. B. Waterhouse, “Dual polarized wide-
to a shielded radiating element which can be suitably used in band aperture stacked patch antennas,” IEEE Transactions on
array configurations, where the rhombic elements providing Antennas and Propagation, vol. 52, no. 8, pp. 2171–2174, 2004.
the two polarizations are stacked on different substrates. In [14] D. Tallini, A. Galli, M. Ciattaglia, L. Infante, A. De Luca, and
this structure, lateral metallic walls prevent mutual-coupling M. Cicolani, “A new low-profile wide-scan phased array for
effects among elements and reduce possible undesired effects UWB applications,” in Proceedings of the European Conference
of power leakage in the substrate. A good input matching on Antennas and Propagation, pp. 1–5, Edinburgh, UK,
is obtained in the whole frequency band required, with November 2007.
very promising performance in terms of radiation pattern [15] A. Galli, G. Valerio, D. Tallini, A. De Luca, and M. Cicolani,
regularity and gain. The basic performance achievable in “Optimization of multifunctional UWB planar phased arrays,”
array configurations has been finally addressed. On this in Proceedings of the European Conference on Antennas and
ground, manufacturing and experimental test of prototypes Propagation, pp. 571–574, Berlin, Germany, March 2009.
of 2D arrays is programmed next. [16] S. Mazzocchi, G. Valerio, M. Zucca, M. Ciattaglia, A. De
Luca, and A. Galli, “New multifunctional UWB planar
antenna based on printed dipoles in rhombic configuration,”
References in Proceedings of the European Conference on Antennas and
[1] B. Allen, M. Dohle, E. Okon, W. Malik, A. Brown, and D. Propagation, Barcelona, Spain, April 2010.
Edwards, Eds., Ultra Wideband Antennas and Propagation for [17] A. A. Qureshi, M. U. Afzal, T. Taqueer, and M. A. Tarar,
Communications, Radar and Imaging, Wiley, 2006. “Performance analysis of FR-4 substrate for high frequency
[2] W. Wiesbeck, G. Adamiuk, and C. Sturm, “Principles of UWB microstrip antennas,” in Proceedings of the China-Japan Joint
antennas basic properties and design,” Proceedings of the IEEE, Microwave Conference (CJMW ’11), pp. 159–162, April 2011.
vol. 97, no. 2, pp. 372–385, 2009. [18] http://www.modefrontier.com.
[19] C. Poloni and V. Pediroda, “GA coupled with computationally

expensive simulations: tools to improve efficiency,” in Algo-
rithms and Evolution Strategies in Engineering and Computer
Science, pp. 267–288, Wiley, 1997.
[20] CST Microwave Studio Manual, CST, Germany, 2002.
doi:10.1155/2012/853651
Research Article
Design and Analysis of Wideband Nonuniform Branch Line
Coupler and Its Application in a Wideband Butler Matrix
Yuli K. Ningsih,1, 2 M. Asvial,1 and E. T. Rahardjo1

1 Antenna Propagation and Microwave Research Group (AMRG), Department of Electrical Engineering,
Universitas Indonesia, New Campus UI, West Java, Depok 16424, Indonesia
2 Department of Electrical Engineering, Trisakti University, Kyai Tapa, Grogol, West Jakarta 11440, Indonesia
Correspondence should be addressed to Yuli K. Ningsih, yuli kn@yahoo.com
Received 10 August 2011; Accepted 2 December 2011
Academic Editor: Tayeb A. Denidni
Copyright © 2012 Yuli K. Ningsih et al. This is an open access article distributed under the Creative Commons Attribution License,
This paper presents a novel wideband nonuniform branch line coupler. An exponential impedance taper is inserted, at the series
arms of the branch line coupler, to enhance the bandwidth. The behavior of the nonuniform coupler was mathematically analyzed,
and its design of scattering matrix was derived. For a return loss better than 10 dB, it achieved 61.1% bandwidth centered
at 9 GHz. Measured coupling magnitudes and phase exhibit good dispersive characteristic. For the 1 dB magnitude difference
and phase error within 3◦ , it achieved 22.2% bandwidth centered at 9 GHz. Furthermore, the novel branch line coupler was
implemented for a wideband crossover. Crossover was constructed by cascading two wideband nonuniform branch line couplers.
These components were employed to design a wideband Butler Matrix working at 9.4 GHz. The measurement results show that
the reflection coefficient between the output ports is better than 18 dB across 8.0 GHz–9.6 GHz, and the overall phase error is less
than 7◦ .
1. Introduction In general, the Butler Matrix is an N × N passive feeding

network, composed of branch line coupler, crossover, and
Recently, a switched-beam antenna system has been widely phase shifter. The bandwidth of the Butler Matrix is greatly
used in numerous applications, such as in mobile commu- dependent on the performance of the components. However,
nication system, satellite system, and modern multifunction the Butler Matrix has a narrow bandwidth characteristic
radar. This is due to the ability of the switched-beam antenna due to branch line coupler and crossover has a limited
to decrease the interference and to improve the quality of bandwidth.
transmission [1, 2] and also to increase gain and diversity As there is an increased demand to provide high data
[3]. throughput [16], it is essential that the Butler Matrix has
The switched-beam system consists of a multibeam to operate over a wide frequency band when used for angle
switching network and antenna array. The principle of a diversity. Therefore, many papers have reported for the
switched-beam is based on feeding a signal into an array of bandwidth enhancement of branch line coupler [17–20]. In
antenna with equal power and phase difference. Different reference [17, 18], design and realization of branch line cou-
structures of multibeam switching networks have been pler on multilayer microstrip structure was reported. These
proposed, such as the Blass Matrix, the Nolen Matrix, the designs can achieve a wideband characteristic. However, the
Rotman Lens, and the Butler Matrix [4]. One of the most disadvantages of these designs are large in dimension and
widely known multibeam switching networks with a linear bulk.
antenna is the Butler Matrix. Indeed, it seems to be the most Reference [19] introduces a compact coupler in an N-
attractive option due to its design simplicity and low power section tandem-connected structure. The design resulted in
loss [5–15]. a wide bandwidth. Another design, two elliptically shaped
A =1
P1 √ P3
Z(z)/ 2
P1 P3 S 11 S 31
4.75
Zo Zo
8 12.75 14.5
S 21 S 41
3
P2 P4
P2 P4
Figure 2: Circuit of the nonuniform branch line coupler.

Figure 1: Geometry structure of a new nonuniform branch line
coupler design with exponential impedance taper at the series arm.
+1/2
1 √ 1
Z(z)/ 2
microstrip lines which are broadside coupled through an
1 1 Se12
elliptically shaped slot, was employed in [20]. This design Se11
was used in a UWB coupler with high return loss and
isolation. However, these designs require a more complex
manufacturing. +1/2 1 1
In this paper, nonuniform branch line coupler using
exponential impedance taper is proposed which can enhance 1 1
bandwidth and can be implemented for Butler Matrix, as Open circuit stub
shown in Figure 1. Moreover, it is a simple design without (2 separate 2-ports)
needs of using multilayer technology. This will lead in cost
(a) Even mode (e)
reduction and in design simplification.
To design the new branch line coupler, firstly, the +1/2
1 √ 1
series arm’s impedance is modified. The shunt arm remains Z(z)/ 2
unchanged. Reduced of the width of the transmission line 1 1 So12
So11
at this arm is desired by modifying the series arm. Next, by
exponential impedance taper at the series arm, a good match
over a high frequency can be achieved.
+1/2 1 1
2. Mathematical Analysis of Nonuniform 1 1

Branch Line Coupler Short circuit stub
(2 separate 2-ports)
The proposed nonuniform branch line coupler use λ/4
(b) Odd mode (o)
branches with impedance of 50 Ω at the shunt arms and
use the exponential impedance taper at the series arms, as Figure 3: Decomposition of the nonuniform branch line coupler
shown in Figure 1. Since branch line coupler has a symmetric into even and odd modes of excitation.
structure, the even-odd mode theory can be employed to
analyze the nonuniform characteristics. The four ports can
be simplified to a two-port problem in which the even
and odd mode signals are fed to two collinear inputs [22]. A branch line coupler has been designed based on the
Figure 2 shows the schematic of circuit the nonuniform theory of small reflection, by the continuously tapered line
branch line coupler. with exponential tapers [23, 24], as indicated in Figure 1,
The circuit of Figure 2 can be decomposed into the where
superposition of an even-mode excitation and an odd-mode

excitation is shown in Figures 3(a) and 3(b). Z0 e−az , for 0 < z ≤ L,
The ABCD matrices of each mode can be expressed Z(z) = (3)
Z0 eaz , for L ≤ z < 2L,
following [22]. In the case of nonuniform branch line
coupler, the matrices for the even and odd modes become:
which determines the constant a as:
A B Z(z) −1 j
= √
C D 2 j −1 , (1)
1 Z
e a= ln L , (4)
L Zo
A B Z(z) 1 j
= √ . (2)
C D o 2 j 1 L = discrete steps length, Zo = Z (0), and ZL = Z (L).
into equal powers in port 3 (P3 ) and port 4 (P4 ). Considering

(1) to (16), a number of properties of the ideal branch line
coupler maybe deduced from the symmetry and unitary
properties of its scattering matrix. If the series and shunt arm
are one-quarter wavelength, by using (11), resulted in S11 =
0.
As both the even and odd modes of S11 are 0, the values
of S11 and S21 are also 0. The magnitude of the signal at the
coupled port is then the same as that of the input port.
Figure 4: Photograph of a proposed nonuniform branch line Calculating (7) and (8) under the same o, the even and
coupler. odd modes of S21 nonuniform branch line coupler will be
expressed as follows in(17)
Z(z)
Useful conversions for two-port network parameters for Se21 = − √ 1 + j ,
the even and odd modes of S11 and S21 can be defined as 2
(17)
follows [22]: Z(z)
So21 = √ 1 − j .
2
(A + B − C − D)e
Se11 = , (5) Based on (13), S31 can be expressed as follows
ΔY e
(A + B − C − D)o 1 Z(z)
So11 = , (6) S31 = − √ 1+ j − 1− j
ΔY o 2 2
2 1 Z(z)
Se21 = , (7) =− √ 1+ j−1+ j (18)
ΔY e 2 2
Z(z)
2 = −j √ .
So21 = , (8) 2
ΔY e
where Following (14), S41 nonuniform branch line coupler can be
calculating as follows
ΔY e = (A + B + C + D)e , (9)
1 Z(z)
S41 = −√ 1+ j + 1− j
ΔY = (A + B + C + D)o .
o
(10) 2 2
1 Z(z)
=− √ 1+ j +1− j (19)
Since the amplitude of the incident waves for these two 2 2
ports are ±1/2, the amplitudes of the emerging wave at each Z(z)
port of the nonuniform branch line coupler can be expressed =− √ .
2
as [22]:
From this result, both S31 and S41 nonuniform branch
1 e
line couplers have equal magnitudes of −3 dB. Therefore, due
S11 = S + So11 , (11)
2 11 to symmetry property, we also have that S11 = S22 = S33 =
1 e S44 = 0, S13 = S31 , S14 = S41 , and S21 = S34 . Therefore, the
S21 = S − So11 , (12) nonuniform branch line coupler has the following scattering
2 11
matrix in (20):
1 e
S31 = S + So21 , (13) ⎡ ⎤
2 21 0 0 j 1
1 e Z(z) ⎢
⎢0 0 1 j
⎥
⎥
S41 = S − So21 . (14) S=− √ ⎢ ⎥. (20)
2 21 2 ⎣ j 1 0 0 ⎦
1 j 0 0
Parameters even and odd modes of S11 nonuniform
branch line coupler can be expressed as (15) and (16) as
follows:
3. Fabrication and Measurement
Result of Wideband Nonuniform Branch
Z(z) −1 + j − j + 1 Line Coupler
Se11 = √ = 0, (15)
2 −1 + j + j − 1
To verify the equation, the nonuniform branch line coupler
Z(z) 1 + j − j − 1 was implemented and its S-parameter was measured. It
So11 = √ = 0. (16) was integrated on TLY substrate, which has a thickness
2 1+ j + j+1
of 1.57 mm. Figure 4 shows a photograph of a wideband
An ideal branch line coupler is designed to have zero nonuniform branch line coupler. Each branch at the series
reflection power and splits the input power in port 1 (P1 ) arm comprises an exponentially tapered microstrip line
0
6 6.5 7 7.5 8 8.5 9 9.5 10 10.5 11 11.5 12 12.5
S-parameters magnitude (dB)
−5
−10
−15
−20
−25
Figure 8: Photograph of microstrip nonuniform crossover.
−30
Frequency (GHz)
S-parameters magnitude (dB)

S (1, 1) S (3, 1)
S (2, 1) S (4, 1) 8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10
−5
Figure 5: Measurement result for nonuniform branch line coupler. −10
−15
−20
200
Phase characteristic (deg)
150 −25
100 −30
50 Frequency (GHz)
0
−50 6 7 8 9 10 11 12 S (1, 1) S (3, 1)
S (2, 1) S (4, 1)
−100
−150 Figure 9: Measurement result for nonuniform crossover.
−200
Frequency (GHz)
S (4, 1)
S (3, 1)
Figure 6: Phase characteristic of nonuniform branch line coupler.

P5 P6 P7 P8
45◦ 45◦ Phase shifter
3 dB hybrid
Cross over coupler
Figure 10: Final layout of the proposed wideband Butler Matrix
4 × 4.
P1 P2 P3 P4
Figure 7: Basic schematic of the 4 × 4 Butler Matrix [21].

the coupling ratio is supposed approximately 3 ± 1 dB, the
bandwidth of about 22.2% centered at 9 GHz.
which transforms the impedance from Zo = 50 ohms to As expected, the phase difference between port 3 (P3 ) and
ZL = 50.6 ohms. This impedance transformation has been port 4 (P4 ) is 90◦ . At 9 GHz, the phases of S31 and S41 are
designed across a discrete step length L = 6.75 mm. 85.54◦ and 171◦ , respectively. These values differ from ideal
Figure 5 shows the measured result frequency response of value by 4.54◦ . The average phase error or phase unbalance
the novel nonuniform branch line coupler. For a return loss between two branch line coupler outputs is about 3◦ . But
and isolation better than 10 dB, it has a bandwidth of about even the phase varies with frequency; the phase difference is
61.1%; it extends from 7 to 12.5 GHz. In this bandwidth, almost constant and very close to ideal value of 90◦ as shown
the coupling ratio varies between 2.6 dB up to 5.1 dB. If in Figure 6.
Frequency(GHz) Frequency (GHz)

0 0
1 2 3 4 5 6 7 8 9 10 11 8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10
−2 −5
Insertion loss (dB)
−10
−4
Return loss (dB)

−15
−6 −20
−8 −25
−30
−10
−35
−12 −40
−45
S (5, 1) simulated S (5, 1) measured
S (8, 1) simulated S (8, 1) measured S (1, 1) simulated S (1, 1) measured
S (7, 1) measured S (6, 1) simulated S (4, 4) simulated S (4, 4) measured
S (7, 1) simulated S (6, 1) measured (a) Input port 1 or 4 are excited
(a) Input port 1 excitation
Frequency (GHz)
Frequency (GHz) 0
0 8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10
8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10 −5
Return loss (dB)

−2 −10
Insertion loss (dB)
−4 −15
−6 −20
−8 −25
−30
−10
−35
−12
S (5, 2) simulated S (5, 2) measured S (3, 3) simulated S (3, 3) measured
S (7, 2) measured S (6, 2) simulated (b) Input port 2 or 3 are excited
Figure 12: Return loss of the proposed Butler Matrix when different
(b) Input port 2 excitations ports are fed.
Frequency (GHz)
0
8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10
−2
Insertion loss (dB)
−4
4. Design and Fabrication of the Wideband
−6
Butler Matrix
−8
−10 Figure 7 shows the basic schematic of the 4 × 4 Butler Matrix
−12 [21]. Crossover also known as 0 dB couplers is a four-port
device and must provide for a very good matching and
isolation, while the transmitted signal should not be affected.
S (7, 3) measured S (6, 3) simulated In order to achieve wideband characteristic crossover, this
S (7, 3) simulated S (6, 3) measured paper proposes the cascade of two nonuniform branch line
couplers.
(c) Input port 3 excitations
Figure 8 shows the microstrip layout of the optimized
Frequency (GHz)
0 crossover. The crossover has a frequency bandwidth of
1 2 3 4 5 6 7 8 9 10 11
−2 1.3 GHz with VSWR = 2, which is about 22.2% of its centre
Insertion loss (dB)
−4 frequency at 9 GHz. Thus, it is clear from these results

−6 that a nonuniform crossover fulfills most of the required
−8
specifications, as shown in Figure 9.
−10
Figure 10 shows the layout of the proposed wideband
Butler Matrix. This matrix uses wideband nonuniform
−12
branch line coupler, wideband nonuniform crossover, and
S (5, 4) simulated S (6, 4) simulated phase-shift transmission lines.
S (5, 4) measured S (6, 4) measured The wideband Butler Matrix was measured using Net-
S (7, 4) simulated S (8, 4) simulated work Analyzer. Figure 11 shows the simulation and measure-
S (7, 4) measured S (8, 4) measured
ment results of insertion loss when a signal was fed into
(d) Input port 4 excitations port 1, port 2, port 3, and port 4, respectively. The insertion
loss are varies between 5 dB up to 10 dB. For the ideal
Figure 11: Insertion loss of the proposed Butler Matrix when
different ports are fed. Butler matrix, it should be better than 6 dB. Imperfection
60 Table 1: Output phase difference and estimated direction of

Phase different (deg)
50 generated beam.
40
30 P5 (◦ ) P7 (◦ ) P6 (◦ ) P8 (◦ ) β (◦ ) θ (◦ )
20 P1 45 90 135 180 45 14.4 [1L]
10 P2 135 0 225 90 −135 −48.6 [2R]
0
8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10 P3 90 225 0 135 135 48.6 [2L]
Frequency (GHz) P4 180 135 90 45 −45 −14.4 [1R]
Phase (port 3)-phase (port 1)
Phase (port 4)-phase (port 2) of fabrication could contribute to reduction of the insertion
(a) Input port 1 excitation loss.
The simulated and measured results of the return loss
Frequency (GHz)
−115 at each port of the wideband 4 × 4 Butler Matrix is shown
8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10

−120 in Figure 12. For a return loss better than 10 dB, it has a
−125 bandwidth about 17% centered at 9.4 GHz.
−130
Figure 13 shows the phase difference of measured results
−135
−140 when a signal was fed into port 1, port 2, port 3, and port 4,
−145 respectively. The overall phase error was less than 7◦ . There
−150 are several possible reasons for this phase error. A lot of bends
in high frequency can produce phase error. Moreover, the
imperfection of soldering, etching, alignment, and fastening
Phase (port 3)-phase (port 1) also could contribute to deviation of the phase error.
Phase (port 2)-phase (port 3) Table 1 shows that each input port was resulted a specific
Phase (port 4)-phase (port 2) linear phase at the output ports. The phase differences each
(b) Input port 2 excitations between the output ports are of the same value. The phase
difference can generate a different beam (θ). If port 1 (P1)
145
140
is excited, the phase difference was 45◦ , the direction of
135 generated beam (θ) will be 14.4◦ for 1L. It is summarized
130 in Table 1.
125
120
115 5. Conclusion
8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10
Frequency (GHz) A novel nonuniform branch line coupler has been
employed to achieve a wideband characteristic by expo-
nential impedance taper technique. It is a simple design
without needs of using multilayer technology and this
will lead to cost reduction and design simplification. The
(c) Input port 3 excitations scattering matrix of the nonuniform branch line coupler
was derived and it was proved that the nonuniform branch
Frequency (GHz) line coupler has equal magnitude of −3 dB. Moreover, the
0 novel nonuniform branch line coupler has been employed
8 8.2 8.4 8.6 8.8 9 9.2 9.4 9.6 9.8 10
to achieve a wideband 0 dB crossover. Furthermore, these
−10
−20
components have been implemented in the Butler Matrix
and that achieves wideband characteristics.
−30
−40
−50 References
−60
[1] T. A. Denidni and T. E. Libar, “Wide band four-port butler
matrix for switched multibeam antenna arrays,” in Proceedings
of the IEEE International Symposium on Personal, Indoor and
Phase (port 6)-phase (port 5) Mobile Radio Communications (PIMRC ’03), vol. 3, pp. 2461–
Phase (port 8)-phase (port 7) 2464, 2003.
Phase (port 7)-phase (port 6) [2] E. Siachalou, E. Vafiadis, S. S. Goudos, T. Samaras, C. S.
(d) Input port 4 excitations Koukourlis, and S. Panas, “On the design of switched-beam
wideband base stations,” IEEE Antennas and Propagation
Figure 13: Phase difference of the proposed Butler Matrix when Magazine, vol. 46, no. 1, pp. 158–167, 2004.
different ports are fed. [3] P. S. Hall and S. J. Vetterlein, “Review of radio frequency
beamforming techniques for scanned and multiple beam
antennas,” IEE Proceedings H, vol. 137, no. 5, pp. 293–303, [20] M. E. Bialkowski, N. Seman, and M. S. Leong, “Design of a
1990. compact ultra wideband 3 db microstrip-slot coupler with
[4] W-D. Wirth, Radar Techniques Using Array Antennas, IEE high return losses and isolation,” in Asia Pacific Microwave
Publishers, Stevenage, UK, 2001. Conference (APMC ’09), pp. 1334–1337, St. Lucia, Australia,
[5] S. Y. Zheng, S. H. Yeung, W. S. Chan, and K. F. Man, “Broad- December 2009.
band butler matrix optimized using jumping genes evolu- [21] R. P. Hecken, “A near-optimum matching section without
tionary algorithm,” in Proceedings of the IEEE International discontinuities,” IEEE Transactions on Microwave Theory and
Conference on Industrial Technology (IEEE ICIT ’08), Hong Techniques, vol. 20, no. 11, pp. 734–739, 1972.
Kong, April 2008. [22] D. M. Pozar, Microwave Engineering, John Wiley& Sons, New
[6] K. Wincza and K. Sachse, “Broadband Butler matrix in mi- York, NY, USA, 2nd edition, 1998.
crostrip multilayer technology designed with the use of [23] M. Bona, L. Manholm, J. P. Starski, and B. Svensson, “Low
three-section directional couplers and phase correction Net- loss compact butler matrix for a microstrip antenna,” IEEE
works,” in Proceedings of the 18th International Conference on Transactions on Microwave Theory and Techniques, vol. 50, no.
Microwave Radar and Wireless Communications(MIKON ’10), 9, pp. 2069–2075, 2002.
Cracow, Poland, June 2010. [24] M. Kobayashi and N. Sawada, “Analysis and synthesis of
[7] A. M. El Tager and M. A. Eleiwa, “Design and implementation tapered microstrip transmission lines,” IEEE Transactions on
of a smart antenna using butler matrix for ISM band,” Microwave Theory and Techniques, vol. 40, no. 8, pp. 1642–
in Proceedings of the Progress in Electromagnetics Research 1646, 1992.
Symposium (PIERS ’09), pp. 571–575, Beijing, China, March
2009.
[8] Y. S. Jeong and T. W. Kim, “Design and analysis of swapped
port coupler and its application in a miniaturized butler
matrix,” IEEE Transactions on Microwave Theory and Tech-
niques, vol. 58, no. 4, pp. 764–770, 2010.
[9] C. Collado, A. Grau, and F. De Flaviis, “Dual-band butler
matrix for WLAN systems,” in IEEE MTT-S International
Microwave Symposium Digest, vol. 2005, pp. 2247–2250, 2009.
[10] K. Wincza, S. Gruszczynski, and K. Sachse, “Integrated four-
beam dual-band antenna array fed by broadband Butler
matrix,” Electronics Letters, vol. 43, no. 1, pp. 7–8, 2007.
[11] T. N. Kaifas and J. N. Sahalos, “On the design of a single-layer
wideband Butler matrix for switched-beam UMTS system
applications,” IEEE Antennas and Propagation Magazine, vol.
48, no. 6, pp. 193–204, 2006.
[12] K. Wincza and S. Gruszczynski, “A broadband 4 × 4 butler
matrix for modern-day antennas,” in Proceedings of the 35th
European Microwave Conference, pp. 1331–1334, Paris, France,
October 2005.
[13] S. Gruszczyński, K. Wincza, and K. Sachse, “Reduced sidelobe
four-beam N-element antenna arrays fed by 4 × 4 N butler
matrices,” IEEE Antennas and Wireless Propagation Letters, vol.
5, no. 1, pp. 430–434, 2006.
[14] Y. C. Su, M. E. Bialkowski, F. C. E. Tsai, and K. H. Cheng,
“UWB switched-beam array antenna employing UWB butler
matrix,” in Proceedings of the IEEE International Workshop on
Antenna Technology: Small Antennas and Novel Metamaterials
(iWAT ’08), pp. 199–202, Hsinchu, Taiwan, March 2008.
[15] J. He, B. Z. Wang, Q. Q. He, Y. X. Xing, and Z. L. Yin,
“Wideband x-band microstrip Butler matrix,” Progress in
Electromagnetics Research, vol. 74, pp. 131–140, 2007.
[16] Y. Liuqing and G. B. Giannakis, “Ultra wideband communi-
cations,” IEEE Signal Processing Magazine, vol. 21, no. 6, pp.
26–54, 2004.
[17] S. Banba and H. Ogawa, “Multilayer MMIC directional coup-
lers using thin dielectric layers,” IEEE Transactions on Mi-
crowave Theory and Techniques, vol. 43, no. 6, pp. 1270–1275,
1995.
[18] J. Sebastien and G. Y. Delisle, “Microstrip EHF butler matrix
design and realization,” ETRI Journal, vol. 27, no. 6, pp. 788–
797, 2005.
[19] J. H. Cho, H. Y. Hwang, and S. W. Yun, “A design of wideband
3-dB coupler with N-section microstrip tandem structure,”
IEEE Microwave and Wireless Components Letters, vol. 15, no.
2, pp. 113–115, 2005.
doi:10.1155/2012/264618
Research Article
Isolation Improvement of a Microstrip Patch Array Antenna for
WCDMA Indoor Repeater Applications
Hongmin Lee and Jinwon Park

Department of Electronic Engineering, Kyonggi University, Suwon 443-760, Republic of Korea
Correspondence should be addressed to Hongmin Lee, hmlee@kyonggi.ac.kr
Received 15 July 2011; Revised 5 November 2011; Accepted 17 November 2011
Academic Editor: Byungje Lee
Copyright © 2012 H. Lee and J. Park. This is an open access article distributed under the Creative Commons Attribution License,
This paper presents the isolation improvement techniques of a microstrip patch array antenna for the indoor wideband code
division multiple access (WCDMA) repeater applications. One approach is to construct the single-feed switchable feed network
structure with an MS/NRI coupled-line coupler in order to reduce the mutual coupling level between antennas. Another approach
is to insert the soft surface unit cells near the edges of the microstrip patch elements in order to reduce backward radiation
waves. In order to further improve the isolation level, the server antenna and donor antenna are installedin orthogonal direction.
The fabricated antenna exhibits a gain over 7 dBi and higher isolation level between server and donor antennas below −70 dB at
WCDMA band.
1. Introduction antennas on a substrate are available. One approach is to con-

struct a periodic structure, such as photonic bandgap (PBG)
Microstrip patch antennas are used in modern communica- or electromagnetic bandgap (EBG) [1, 2], surrounding the
tion systems due to their low cost, lightweight, and planar patch antenna. Another approach is to use the concept of
structure. However, a microstrip patch antenna has some artificial soft and hard surfaces [3]. These surfaces have led
disadvantages. to a wide range of applications in antennas other microwave
In the design of a microstrip patch array antenna, the systems [4]. Both EBG and soft surfaces can be used to
radiating patch elements have to be placed in close prox- suppress surface wave propagation. The main difference is
imity. It produces significant mutual coupling effect that that soft surfaces exhibit bandgaps in only one direction,
deteriorates the antenna performance. The simplest solution but they offer the best performance in most applications
to improve the antenna isolation is to increase the physical of antennas. However, the conventional soft surfaces are
separation between the antennas. Although the isolation made on the same ground plane of a patch antenna and
of the antennas becomes higher as the separation length the presence of via near the patch may raise the resonant
is increased, the size of the antenna will subsequently be frequency of the patch if the soft surface strip or cell is
larger. However, the size of the antenna ground plane is very close to the patch. On the other hand, it requires a
limited in the practical microstrip patch antenna design. considerable area to form a bandgap structure.
Particularly when the patch antenna is printed on high The purpose of this paper is to present a new method for
dielectric substrates, its radiation pattern is considerably enhancing an isolation of a microstrip patch array antenna,
affected by surface waves. On a finite ground plane, surface whose Tx/Rx band is located very near. In order to reduce
waves propagate until they reach an edge where they are the mutual coupling between antennas that are located in the
reflected back and diffracted by the edges. These cause same plane, a passive switching feed network is used [5]. A
a significant amount of wasted power in the backward new isolated soft surfaces structure that does not share the
hemisphere below the ground plane. Various solutions for ground plane of a patch is also used for the reduction of the
the improvement of the radiation performance of patch surface wave. The proposed new soft surface consists of a
number of metal mushroom-type structures that is located l1 Port 2

near the edges of substrate, and it is not connected with a Z1 ZL1
ground plane [6, 7]. A two-by-two microstrip patch array
Port 1 (Output)
antenna for the indoor wideband code division multiple
access (WCDMA) repeater applications is designed and Z0 Zin 1
tested experimentally. l2 Port 3
(Input)
Z2 ZL2
2. Switchable Feed Network for Antenna Zin (Output)
The schematic of a single-feed switchable feed network Zin 2
is shown in Figure 1. It basically consists of two quarter-
wavelength branch lines with characteristic impedances of Figure 1: Schematic of a single-feed switchable feed network.
Z1 and Z2 lengths of l1 and l2 . Two different rectangular
microstrip patch antennas of different sizes are connected to 11.5 mm
output ports 2 and 3. The resonant frequencies of these two
antennas are f1 and f2 , and the input impedances of each Port 4 0.85 mm Port 3
antenna are ZL1 and ZL2 . Lengths of l1 and l2 are chosen as
a quarter wavelength of λ1 /4 and λ2 /4, respectively. When a Port 2 Port 1
signal with a frequency of f = f1 is applied to port 1, the
input impedance of the upper branch Zin1 is
Z12 15.7 mm
Zin1 = . (1)
ZL1 Figure 2: Geometry of 3 dB MS/NRI coupled-line coupler.
If the resonant frequencies of two patches are very closely
located ( f1 ≈ f2 ), the two quarter wavelength branch line Table 1: The operation of the single-feed switchable feed network.
lengths become similar (l1 ≈ l2 ). Since the patch antenna
connected to port 3 is off-resonant, the input impedance of f [GHz] Port Zin ZL Switch state
the antenna is almost reactive (ZL2 ≈ jX). Hence, the input Port 2 50 Ω 50 Ω “On”
f = f1
impedance of the lower branch Zin is Port 3 ≈ ∞Ω jx ≈ 0 Ω “Off ”
Port 2 ≈ ∞Ω jx ≈ 0 Ω “Off ”
ZL2 + jZ2 tan β2 l2 Z2 f = f2
Zin2 = Z2 ≈ −j 2 . (2) Port 3 50 Ω 50 Ω “On”
Z2 + jZL2 tan β2 l2 x
When the patch antenna was off-resonant, its input reactance
neared zero (x → 0). Using x → 0 in (2), the input 3 (|S21 | = |S31 |). Usually the total length of the coupler
impedance of port 2 neared infinity (Zin2 → ∞). Therefore, and the spacing between the lines should be adjusted until a
it acts as an open circuit. As a result, the patch at port 2 3 dB coupling level is acquired. A 3 dB MS/NRI coupled-line
would be at resonance with the patch at the off-resonant coupler is designed on a Rogers RO3210 substrate (relative
port 3. If a signal with a frequency of f = f2 is applied dielectric constant = 10.2, thickness = 2.54 mm).
to port 1, the patch at port 3 will be at resonance with the The NRI transmission line was 15.7 mm long and the
off-resonant patch at port 2. This single-feed switchable feed MS-line was 11.5 mm long. The spacing between MS/NRI
network combines two ports of a microstrip antenna with lines is 1 mm. Figure 3 presents the simulated scattering
a quarter wavelength feed line and each resonant frequency parameter results for the MS/NRI coupled-line coupler of
of the antennas very closely located; it acts as an ideal Figure 2. Quasi-3 dB backward coupling is achieved over
single-pole double-throw (SPDT) switch. The operation the range from 1.7 to 2.7 GHz. The magnitudes of the
of the single-feed switchable feed network when Z1 = simulated S-parameter are listed in Table 2. Figure 4 shows
Z2 = ZL1 = ZL2 = 50 ohm is presented in Table 1. In the geometry of the feed network for two-by-two microstrip
order to further enhance the isolation between the ports, patch array antenna. It is fed by a 50 Ω coaxial probe and
microstrip/negative-refractive-index (MS/NRI) coupled line designed on a Rogers RO3210 substrate to allow the feeding
coupler is used with the switchable feed network. An MS/NRI of a four-element microstrip patch antenna array. The overall
coupler has advantages, compared to conventional edge- dimension of the ground plane is 120 mm × 120 mm. The
coupled microstrip couplers in terms of coupled power and aimed two switching frequencies of the proposed switchable
port isolation [8]. Figure 2 shows the geometry of the 3 dB feed network are 1.95 and 2.14 GHz, respectively.
MS/NRI coupled-line coupler for WCDMA frequency band These frequencies are the center of the Tx and Rx band
(1.92–2.17 GHz) applications. It consists of a conventional in the WCDMA system. In Figure 4, all the ends of each
MS right-handed transmission line edge-coupled with a port of the proposed feed network for two-by-two microstrip
NRI transmission line used in its left-handed range. A patch antenna are made with 50 Ω microstrip lines. The
3 dB implementation of the MS/NRI coupled-line coupler characteristic impedances of an NRI transmission line and
requires the equality of the magnitude at port 2 and port the MS line were Z1 = 90 Ω and Z2 = 75 Ω, respectively.
(dB) 120 mm
0
−10
−20 4 5
S-parameter
−30
−40
120 mm
1
−50
−60
−70
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 3 2
Frequency (GHz)
S11 S31
S21 S41
Figure 3: Simulated S-parameter results for the MS/NRI coupled- Figure 4: Geometry of the feed network for two-by-two microstrip
line coupler of Figure 2. patch array antenna.
Table 2: Simulated S-parameters for the MS/NRI coupled-line 1.27 mm

coupler. 2.54 mm
S-parameter 1.95 GHz [dB] 2.14 GHz [dB]

S11 −12.5 −13.4
Tx Rx
S21 −3.3 −3.6
S31 −3.6 −3.1
−27.5 −32.5 30 mm
120 mm
S41 34 mm
Table 3: Simulated S-parameters of the feed network for a four-

element microstrip patch array.
Rx Tx
Frequency
1.94 2.15
[GHz]
Port isolation [dB] y
Without With Without With
S-parameter MS/NRI MS/NRI MS/NRI MS/NRI
z x 120 mm 1 mm
coupler coupler coupler coupler
S23 −15.18 −18.81 −11.18 −18.78 Figure 5: Geometry of the proposed switchable feed network for
S24 −14.66 −24.73 −12.20 −17.12 two-by-two microstrip patch array antenna.
S25 −7.03 −23.25 −7.8 −19.57
S32 −15.18 −19.81 −11.18 −23.09
S34 −8.50 −24.01 −8.04 −30.99 switched feeder network and a signal was excited at port 1
S35 −16.88 −26.69 −10.43 −18.91 as shown in Figure 4. When the antennas connected at port
S42 −14.66 −24.73 −12.20 −17.10 2 and port 4 were in an on state, the antennas connected at
port 3 and port 5 were in an off state. However, the proposed
S43 −8.50 −24.01 −8.04 −30.99
switchable feed network shown in Figure 1 can operate when
S45 −16.31 −18.26 −11.67 −20.54
the antennas are connected at each port.
S52 −7.03 −23.25 −7.8 −19.57 For the simulation of the port isolation, all the ends
S53 −16.88 −26.69 −10.43 −18.92 of each port were terminated with microstrip patch anten-
S54 −16.04 −18.26 −11.67 −20.54 nas. The simulated scattering parameters of the proposed
switchable feed network compared to those of the switchable
feed network with/without MS/NRI coupled-line coupler are
In order to match these impedances with 50 Ω port, a single listed in Table 3. The switchable feed network with MS/NRI
section of quarter wavelength impedance transformer was coupled-line coupler exhibits much higher isolation than
added. The proposed feed network was designed by using the switchable feed network without MS/NRI coupled-line
(a) 1.95 GHz (b) 2.15 GHz
Figure 6: Simulated surface current distributions.
180 mm
180 mm
19 mm
20 mm
19 mm
(a) Perspective view (b) Side view
Figure 7: Geometry of the proposed indoor WCDMA repeater antenna.
coupler. When the MS/NRI coupled-line coupler was not The antenna consists of three-layer structure: an air layer
used, the scattering parameters (S25 , S34 , S52 , S43 ) between having a thickness of 1 mm, a dielectric substrate layer,
two antennas showed lower isolation. Since the patch and a stacked microstrip patch layer. The square microstrip
antennas were designed to be excited at their fundamental patches are fed by microstrip lines from perpendicular
resonant mode (TM10 ), strong coupling is produced between directions using the proximity coupled method. This switch-
two antennas that are placed parallel to the radiating edge able feed network was designed to achieve beam pattern
direction. It can be seen that the minimum isolation levels reconfigurable array antenna, which generates ±45◦ linearly
between the output ports exceed −18 dB at the aimed polarized slanted beam patterns at the Tx/Rx frequency
two switching frequencies of the proposed switchable feed bands. In order to dual slant beam, two Rx and Tx microstrip
network. patch elements were placed orthogonally. Figure 6 shows
the simulated surface current distribution at the resonant
3. Antenna Design frequency of 1.95 GHz (Tx) and 2.15 GHz (Rx). At the
frequency of 1.95 GHz, two patch antennas for Tx band
Figure 5 shows the geometry of the two-by-two microstrip are resonant, and two patch antennas for Rx-band are off-
patch array antenna using the proposed switchable feed resonant. As a result, most of the surface currents flow
network and proximity coupled square microstrip patch through the feed line for Tx-band antennas. On the other
elements. hand, at the frequency of 2.15 GHz, two patch antennas for
20 mm 3 mm
12 mm l 3 mm
h
w
1.27 mm
(a) Parallel plate waveguide model with a cell (b) Geometry of unit cell
(dB)
−30
−40
−50
S-parameter
−60
−70
−80
1.9 1.95 2 2.05 2.1 2.15 2.2
Frequency (GHz)
w = 6.6 mm w = 7.2 mm
w = 6.8 mm w = 7.4 mm
w = 7 mm
(c) The simulated transmission characteristics with different sizes of unit

cells
Figure 8: Parallel plate waveguide model with different sizes of unit cells.
Rx band are resonant. The perspective view of a microstrip It depends on the physical size of the unit cell However,
patch array antenna system for the indoor WCDMA repeater the bandwidth of each stop band shows very narrow
is shown in Figure 7. The proposed repeater antenna consists characteristics due to the resonant nature of the unit cell
of a server antenna, a donor antenna, and alumina housing. and a high dielectric constant of the substrate. In order to
It occupies a volume of 180 mm × 180 mm × 20 mm. cover the bandwidth within the WCDMA frequency band,
In order to reduce the surface waves radiation from a three different sizes of unit cell array configuration structure
server antenna and a donor antenna, corner-edged via (w = l = 6.6 mm, 7.0 mm, 7.4 mm) were used in this work.
mushroom-type unit cells are formed near the edges of the In addition, a server antenna and a donor antenna backed by
upper dielectric substrate. Figure 8 shows the parallel plate the alumina housings are arranged in orthogonal direction
waveguide model with different sizes of unit cells and the in order to get higher isolation between two antennas.
results of the simulated transmission characteristics as a
function of frequency. As shown in Figure 8(b), the unit cell
consists of two parallel rectangular plates with the same size 4. Experimental Result
(w = l) and a corner-edged via. The height of the via is
1.27 mm. Inside two rectangular plates, dielectric material The photographs of the fabricated two-by-two microstrip
(relative dielectric constant = 10.2, thickness = 1.27 mm) is patch array antenna structure are shown in Figure 9. The
placed. The transmission coefficient S21 of the parallel plate switchable feeder layer and a stacked microstrip patch layer
waveguide ports without cell exhibits near −43 dB, as shown are etched on a Rogers RO3210 substrate (relative dielectric
in Figure 8(c). When the unit cell is inserted between two constant = 10.2) having different thickness of 2.54 mm
parallel plate waveguides, a stop band occurs at a certain fre- and 1.27 mm, respectively. The characteristic impedance
quency. of each of the two branch feed lines from the coaxial
(a) Top view (b) Switchable feed network
(c) Ground plane (d) Antenna with housing
Figure 9: Photographs of the fabricated two-by-two microstrip patch array antenna.
probe was 100 Ω and the input impedance of each of the

0 patch elements with a proximity coupled microstrip feed is
50 Ω. A comparison of the measured return loss (S11 ) and
−10 isolation (S21 ) between a server antenna and a donor antenna
with/without isolated soft surface unit cells is shown in
S-parameter (dB)
−20 Figure 10. It is noted that the resonant frequencies are barely
changed.
−40 In the higher and lower bands, the measured −10 dB
return loss bandwidths are about 84 MHz and 96 MHz,
−50 respectively. It meets the bandwidth requirement for
WCDMA (1.92–2.17 GHz) applications. Compared to the
−60
fabricated antenna without isolated soft surface unit cells,
the fabricated antenna with isolated soft surface unit cells
1.5 2 2.5 exhibits higher isolation level. When the isolated soft surface
Frequency (GHz) unit cells are used, the maximum isolation level at the
frequency of 1.94 and 2.15 GHz is −92 dB and −70 dB,
S11 W/O cell S21 W/O cell
respectively. The measured far-field radiation patterns in
S11 with cell S21 with cell
the x-y plane (θ = 0◦ ) and y-z plane (φ = 0◦ ) at the
Figure 10: Measured return loss and isolation characteristics of frequency of 1.94 and 2.15 GHz are shown, respectively, in
two-by-two microstrip patch array antenna. Figure 11. It shows linear polarized radiation patterns, and
the main direction of the radiated power was changed due
0 0
10 10
5 330 30 5 330 30
0 0
−5 −5
−10 300 60 −10 300 60
−15 −15
−20 −20
−25 −25
−30 270 90 −30 270 90
−25 −25
−20 −20
−15 −15
−10 120 −10 120
240 240
−5 −5
0 0
5 5
210 150 210 150
10 10
180 180
(a) x-y plane (1.94 GHz) (b) y-z plane (1.94 GHz)
0 0
10 10
330 30 330 30
5 5
0 0
−5
−5 300 300
60 −10 60
−10
−15 −15
−20 −20
−25 −25
−30 270 90 −30 270 90
−25 −25
−20 −20
−15
−15
−10
240 120 −10 240 120
−5
−5
0 0
5 5
210 150 210 150
10 10
180 180
Co-pol Co-pol
Cross-pol Cross-pol
Power sum Power sum
(c) x-y plane (2.15 GHz) (d) y-z plane (2.15 GHz)
Figure 11: Measured far-field radiation patterns of the fabricated antenna.
to the placement of two patches. The main beam direction switchable feed network structure with MS/NRI coupled-
was slanted about −45◦ for the lower band (Tx) and +45◦ line coupler for higher isolation level and (2) the isolated
for the higher band (Rx). The measured peak gain/radiation soft surface unit cells structure for reducing the surface
efficiency for Tx/Rx band was 7.1 dBi/77% and 8.9 dBi/80%, waves. Both structures have been discussed in the paper
respectively. through proper numerical simulation. In order to improve
the isolation further, the server antenna and donor antenna
for an indoor repeater system were placed orthogonally.
5. Conclusion As a result, the fabricated server and donor antennas have
small separation of 20 mm and exhibit higher isolation level.
The new techniques for the isolation improvement of a Experimental results shows that the maximum isolation level
microstrip patch array antenna have been presented. The at the frequency of 1.94 and 2.15 GHz is −92 dB and −70 dB,
two main techniques presented here are (1) the single-feed respectively. The proposed techniques can be easily used for
the design of the microstrip patch array antenna with higher

isolation level.
Acknowledgment
This research was supported by the Basic Science Research
Program through the National Research Foundation of Korea
(NRF) funded by the Ministry of Education, Science and
Technology (no. 2010-0011646).
References
[1] Y. J. Park, A. Herschlein, and W. Wiesbeck, “A photonic
bandgap (PBG) structure for guiding and suppressing surface
waves in millimeter-wave antennas,” IEEE Transactions on
Microwave Theory and Techniques, vol. 49, no. 10, pp. 1854–
1859, 2001.
[2] D. Sievenpiper, L. Zhang, R. F. J. Broas, N. G. Alexöpolous, and
E. Yablonovitch, “High-impedance electromagnetic surfaces
with a forbidden frequency band,” IEEE Transactions on
2074, 1999.
[3] P. S. Kildal and A. Kishk, “EM modeling of surfaces with stop or
go characteristics—artificial magnetic conductors and soft and
hard surfaces,” Applied Computational Electromagnetics Society
Journal, vol. 18, no. 1, pp. 32–40, 2003.
[4] P. S. Kildal, A. A. Kishk, and S. Maci, “Special issue on artificial
magnetic conductors, soft/hard surfaces, and other complex
surfaces,” IEEE Transactions on Antennas and Propagation, vol.
53, no. 1, part 1, pp. 2–7, 2005.
[5] H. M. Lee, “Pattern reconfigurable microstrip patch array
antenna using switchable feed-network,” in Proceedings of the
Asia-Pacific Microwave Conference (APMC ’10), pp. 2017–2020,
December 2010.
[6] H. M. Lee and J. K. Kim, “Front-to-back ratio improvement
of a microstrip patch antenna using an isolated soft surface
structure,” in Proceedings of the European Microwave Conference
(EuMC ’09), pp. 385–388, October 2009.
[7] J. H. Kim and H. M. Lee, “Backward wave reduction of a
microstrip patch antenna using dual-band isolated soft surface
structures,” in Proceedings of the IEEE International Symposium
on Antennas and Propagation Society (AP-S ’10), pp. 1–4, July
2010.
[8] R. Islam and G. V. Eleftheriades, “A planar metamaterial co-
directional coupler that couples power backwards,” in Proceed-
ings of the IEEE MTT-S International Microwave Symposium
Digest, vol. 1, pp. 321–324, June 2003.
doi:10.1155/2012/681431
Research Article
Series-Fed Microstrip Array Antenna with Circular Polarization
Tuan-Yung Han
Department of Computer and Communication Engineering, Chienkuo Technology University, Chang-Hua City 500, Taiwan
Correspondence should be addressed to Tuan-Yung Han, arthurhan@ctu.edu.tw
Received 21 July 2011; Revised 3 November 2011; Accepted 5 November 2011
Academic Editor: Miguel Ferrando
Copyright © 2012 Tuan-Yung Han. This is an open access article distributed under the Creative Commons Attribution License,
This study proposes a novel 2 × 2 array antenna design with broadband and circularly-polarized (CP) operation. The proposed
design uses a simple series-fed network to increase the CP bandwidth without requiring one-by-one adjustment of each array
element or a complex feed network. Selecting the appropriate spacing between each array element allows the proposed array
antenna to generate CP radiation with a low axial ratio. Experimental results based on a prototype show that this 2 × 2 microstrip
array antenna achieves a wide 3 dB axial ratio bandwidth of more than 10%. Simulated data are also provided to confirm the
measured results.
1. Introduction than that for the CP array element type. This is because
the process of adjusting the CP performance of each array
Due to their attractive features, such as low profile, light element can be avoided. However, the resulting compound
weight, and ease of manufacturing using printed circuit tech- array gain is 3 dB lower than the array using the CP element.
niques, microstrip array antennas have come into high Sequential rotation of radiating elements can increase the
demand for satellite communication applications. To gen- input impedance bandwidth and polarization purity and
erate circularly-polarized (CP) radiation from a microstrip achieve a good symmetrical radiation pattern [6]. However, a
array antenna, previous research recommends a simple sequential rotation feed network must provide a delay line to
design method that uses a corporate-fed network [1] to excite allow various feed point with different phases to connect to
each array element simultaneously. To allow each array ele- each array element. This results in a relatively complicated
ment to exhibit equal power amplitude and phase distribu- circuit layout compared to the traditional corporate- or
tion, most corporate-fed networks employ transmission lines series-fed network. Several studies indicate that a corporate-
(of the same length) and power dividers. However, the CP fed array antenna using sequential rotation techniques [7, 8]
bandwidth of an array antenna design with a corporate- can improve the bandwidth of a corporate-fed array an-
fed network is usually limited to that of a single array ele- tenna and enhance the purity of polarization. However, dis-
ment. Thus, previous studies propose the method of apply- advantages such as a complex feed network still exist, creating
ing sequential rotation techniques to the feed network to the possibility of producing multiple reflections between the
improve the CP bandwidth of an array antenna effectively element and the feed.
[2–4]. Due to the sequentially rotated structure of the feed This study proposes a novel 2 × 2 CP microstrip array
network (with a single feed point), the phases of the four antenna design using series-fed lines. Experimental results
array elements (2 × 2 array) are usually orientated sequen- show that the proposed array design and one using sequen-
tially at 0◦ , 90◦ , 180◦ , and 270◦ . This offers an impedance tial rotation-fed network achieve similar levels of CP perfor-
bandwidth approximately three times wider than that of a mances, including CP bandwidth and peak gain. Since the
single array element. In this design, the designated array proposed array is fed by a simple series-fed network, there
element can be either linearly polarized (LP) [5] or CP type. is no need for an elaborate phase shifting or power dividing
If an LP array element is used, the design process of the array circuit. This study presents both simulated and measured re-
antenna (with sequential rotation technique) will be easier sults for the proposed design.
Unit: mm
x
Circular patch
16
1.6 mm FR4 5 Foam
Ground plane
d
11 0◦ 90◦
10.5
Ring slot
13.5 1
3
80 150
3 11
13.5 Open stub

1
270◦ 180◦
80 150
(a) (b)
Figure 1: Geometry of the studied microstrip antennas. (a) Single ring-slot-coupled microstrip antenna. (b) 2 × 2 CP microstrip array
antenna.
2. Array Element Structure were generated using IE3D software. Experimental results
indicate that the CP operating bandwidth of the prototype
Figure 1(a) illustrates the geometry of a single aperture- (also defined as 3 dB axial ratio) is approximately 150 MHz
coupled microstrip antenna. Two orthogonal modes of the (5.1%) with respect to the center frequency measured at
antenna can be excited in series by a microstrip feed line 2950 MHz. Since the prototype is microstrip-fed, it can be
through the coupling of the annular-ring slot in the ground arranged in an array element manner.
plane. Choosing appropriate length (optimum at 11 mm) for
the open stub in the feed line allows the antenna to generate 3. New Array Design and Results
good CP radiation. A previous study analyzes the parameters
and design procedure of this aperture-coupled microstrip Figure 1(b) depicts the proposed CP microstrip array anten-
antenna [9]. A prototype was first fabricated according to na. Four ring-slot-coupled microstrip antenna elements were
the dimensions revealed in Figure 1(a). Figure 2 shows the excited through a series-fed network. Simulation results sug-
measured results along with the simulated results, which gest that the element spacing, d, is the dominant parameter
0 10
9
5
8
10 7
Axial ratio (dB)

Return loss (dB)
15 6
5
20 4
25 3
2
30 1
35 0
2.6 2.7 2.8 2.9 3 3.1 3.2 3.3
40 Frequency (GHz)
2.6 2.7 2.8 2.9 3 3.1 3.2 3.3
Frequency (GHz) measured
simulated
(a) (b)
Figure 2: Measured and simulated results for the single CP ring-slot-coupled microstrip antenna. (a) Return loss. (b) Axial ratio.
90◦
180◦
Feed 0◦ Feed
point point
270◦
(a) (b)
Figure 3: Structure of the reference antennas. (a) Corporate feed network, Array B. (b) Sequential rotation feed networks, Array C.
to determine the CP performance of the proposed array, and totypes. Figure also presents simulated results for Array A,
the optimum value is approximately 73 mm for the studied indicating that they agree with the experimental results.
structure (d ∼ 0.76λ0 , where λ0 is referred to the center fre- These results confirm that all of the tested array prototypes
quency of the CP operating bandwidth). To confirm the have good impedance matching within their CP operating
simulated results, a prototype (Array A) was constructed bandwidths. In addition, the CP bandwidth (4.7%) of Array
based on the dimensions presented in Figure 1(b). Two other B is almost the same as that of its array element, and the
prototypes, Arrays B and C, using the corporate and sequen- CP bandwidth (13.8%) of Array C is about three times that
tial rotation feed networks, respectively, were also con- of Array B. As for Array A, the CP bandwidth centered at
structed as references (Figure 3). Except for different feeding 3110 MHz is approximately 11.5% which is slightly less than
networks, the three array prototypes were designed with the that of Array C.
same structure, antenna dimensions, and element spacing. Figure 5 plots the measured radiation patterns of Array
Due to the simplicity of series-fed network design, antenna A at 3110 MHz, revealing good left-hand CP (LHCP) radia-
engineers can ignore the process of adjusting the CP char- tions in the broadside direction. The main beam tilts slightly
acteristics of each array element. However, the complex feed (about 3 degrees) to the left side in the x-z plane. This might
network designs in sequential rotational-fed and corporate- be because the array elements are not fed with an exactly
fed array antennas necessitate tuning the CP performance of equal power level. Nevertheless, the measured peak gain of
each array element. Array A is approximately 12 dBi, which is only 0.3 dB lower
Figures 4(a) and 4(b) present the measured return loss than Array C. This small difference might be due to the feed-
and axial ratio against frequency for the respective array pro- ing phase errors of each element. Table 1 summarizes the ex-
10
0
9
5
8
10 7
Return loss (dB)
Axial ratio (dB)

15 6
20 5
25 4
3
30
2
35
1
40
2.6 2.7 2.8 2.9 3 3.1 3.2 3.3 0
2.6 2.7 2.8 2.9 3 3.1 3.2 3.3
Frequency (GHz)
Frequency (GHz)
Array A Array C Array A Array C

Array B Array A, simulated Array B Array A, simulated
(a) (b)
Figure 4: Measured results for the three array prototypes; d = 73 mm. (a) Return loss. (b) Axial ratio.
20 20
Magnitude (dB)
Magnitude (dB)
10 10
0 0
−10 −10
−20 −20
−30 −30
−40 −40
180 150 120 90 60 30 0 −30 −60 −90 −120 −150 −180 180 150 120 90 60 30 0 −30 −60 −90 −120 −150 −180
Angle (deg) Angle (deg)
LHCP
RHCP
(a) (b)
Figure 5: Radiation patterns of Array A measured at 3110 MHz. (a) x-z plane. (b) y-z plane.
Table 1: Experimental results for the studied single element and 4. Conclusions
array prototypes.
This study presents a 2 × 2 circularly polarized microstrip
3 dB axial-ratio CP center Peak gain array antenna using a series-fed network. Experimental re-
bandwidth (MHz, %) frequency (MHz) (dBi) sults indicate that this array has a broad CP operating band-
Single element 150, 5.1 2950 7 width and an acceptable antenna gain. Moreover, the pro-
Array A 360, 11.5 3110 12 posed design is relatively simple compared to the traditional
Array B 140, 4.7 2980 12.4 array antenna, which uses a corporate or sequential rotation
Array C 410, 13.8 2965 12.3 feed network.
References
[1] P. S. Hall and C. M. Hall, “Coplanar corporate feed effects in
perimental results of CP performance for all three array pro- microstrip patch array design,” IEE Proceedings H, vol. 135, no.
totypes. These results show that the CP bandwidth of Array 3, pp. 180–186, 1988.
A is approximately 6.8% larger than that of Array B, and [2] J. W. Baik, K. J. Lee, W. S. Yoon, T. H. Lee, and Y. S. Kim,
2.3% smaller than that of Array C. Although array A demon- “Circularly polarised printed crossed dipole antennas with
strates a slightly narrower bandwidth than array C, its peak broadband axial ratio,” Electronics Letters, vol. 44, no. 13, pp.
785–786, 2008.
gain measured in the boresight direction is almost similar [3] M. Elhefnawy and W. Ismail, “A microstrip antenna array for
(compared to Arrays B and C) at approximately 12 dBi. Array indoor wireless dynamic environments,” IEEE Transactions on
A also possesses a very simple feed network structure that Antennas and Propagation, vol. 57, no. 12, pp. 3998–4002, 2009.
does not require other circuit features, such as phase shifting [4] R. Caso, A. Buffi, M. Rodriguez Pino, P. Nepa, and G.
or power dividing circuits. Thus, Array A is a preferred can- Manara, “A novel dual-feed slot-coupling feeding technique for
didate compared to Array C if a simple feed network struc- circularly polarized patch arrays,” IEEE Antennas and Wireless
ture is required. Propagation Letters, vol. 9, pp. 183–186, 2010.
[5] J. Huang, “A technique for an array to generate circular polar-

ization with linearly polarized elements,” IEEE Transactions on
Antennas and Propagation, vol. AP-34, pp. 1113–1124, 1986.
[6] H. Evans, P. Gale, and A. Sambell, “Performance of 4 × 4
sequentially rotated patch antenna array using series feed,”
[7] D. C. Chung, S. Y. Choi, Y. H. Ko, J. H. Lee, and M. H.
Kwak, “Circularly polarized HTS microstrip antenna array,”
IEEE Transactions on Applied Superconductivity, vol. 13, no. 2,
pp. 301–304, 2003.
[8] M. N. Jazi and M. N. Azarmanesh, “Design and implementa-
tion of circularly polarised microstrip antenna array using a
new serial feed sequentially rotated technique,” IEE Proceedings:
Microwaves, Antennas and Propagation, vol. 153, no. 2, pp. 133–
140, 2006.
[9] J. S. Row, “Design of aperture-coupled annular-ring microstrip
antennas for circular polarization,” IEEE Transactions on Anten-
nas and Propagation, vol. 53, no. 5, pp. 1779–1784, 2005.
doi:10.1155/2012/980252
Research Article
Vertical Meandering Approach for Antenna Size Reduction
Li Deng,1, 2 Shu-Fang Li,1 Ka-Leung Lau,2 and Quan Xue2, 3

1 Key Laboratory of Universal Wireless Communication, Beijing University of Posts and Telecommunications,
Beijing 100876, China
2 Shenzhen Research Institute, City University of Hong Kong, Shenzhen 518057, China
3 State Key Laboratory of Millimeter Waves, City University of Hong Kong, Kowloon, Hong Kong
Correspondence should be addressed to Li Deng, kingdl2001@163.com
Received 15 July 2011; Accepted 30 September 2011
Academic Editor: Ahmed A. Kishk
Copyright © 2012 Li Deng et al. This is an open access article distributed under the Creative Commons Attribution License, which
permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
A novel vertical meandering technique to reduce the lateral size of a planar printed antenna is presented. It is implemented by
dividing a conventional spiral patch into a different number of segments and placing them on different sides of the microwave
substrate with vias as the connections. To confirm the validity of this technique, measured electrical performance and radiation
characteristics of five antennas with different numbers of segments are compared. The smallest antenna is reduced in size by 84%
when compared with the conventional printed spiral antenna.
1. Introduction other in the far-field region. This cancellation renders a

considerable portion of opposing currents ineffective as far
The topic of antenna miniaturization has been a subject of as radiation efficiency is concerned and leads to a very low
interest for more than half a century, but in recent years, it radiation resistance that might have increased using a two-
has attained significant attention because of an exorbitant strip meandered line [8]. Consequently, these antennas are
demand for mobile wireless communication systems. The difficult to match, and yet require a very low temperature of
need for antenna miniaturization stems from the fact that operation to control material losses [9].
most mobile platforms have a limited space for all of The technique for lengthening the current path men-
the required antennas in ever-increasing wireless systems. tioned above is almost based on a coplanar or single-layer
Miniature antennas are in high demand, since the antenna microstrip structure. The current path lengthening for a
size often imposes a significant limitation on the overall size fixed patch antenna can also be obtained by using vertical
of a portable wireless system. meandering technique. By vertical meandering, we can use
An example of a miniaturized antenna is a meandered the area of substrate more efficiently. Furthermore, we can
antenna where a half wavelength dipole is made compact by reduce the cancellation in the above-mentioned conventional
meandering the wire [1]. A similar approach can be applied meandering techniques effectively by rational design of the
to design a meander-type slot antenna [2]. Meandering vertical meandering lines.
the excited patch surface current paths in the microstrip In this paper, a novel vertical meandering technique
antenna’s radiating patch is also an effective method for aimed at reducing the size of a printed spiral patch antenna
achieving a lowered fundamental resonant frequency for the is proposed. By utilizing this technique on a single turn
microstrip antenna [3–7]. On the other hand, meandered conventional printed spiral patch antenna, its resonant
antennas are very hard to match to a 50 Ohm line. This frequency can maximally drop from 1.53 GHz to 0.61 GHz,
difficulty is due to the fact that the radiation of almost which is 60%. Therefore, the antenna size can be reduced by
in-phase electric currents flowing in opposite directions 84%. The antenna finds a reasonable impedance bandwidth
on closely spaced wires or patches tends to cancel each (S11 ≤ −10 dB) of 2.6%.
Via
Spiral-shaped
patch
w
w
a a
r2 r1 r1 y
Ground plane y r2 x
Ground plane
Feeding probe
x Feeding probe
L1 50 Ohm microstrip line L1
L2
L2
b
b
Bottom side Bottom side
Top side Top side
(a) (b)
Via Top side
Bottom side
z
y
x
Feeding probe
Ground plane
(c)
Figure 1: (a) Top view of the conventional printed spiral antenna. (b) Top view of the novel printed spiral antenna. (c) 3D view of the novel
structure.
2. Antenna Structure the microwave substrate. Finally, vertical conducting vias are
used to connect the corners of the segments on the top
Figure 1(a) shows the geometry of a conventional printed and bottom layers together (as is well known, the vias are
spiral antenna. It consists of a small rectangular ground plane used widely in PCB for connections between layers, it is easy
etched on the bottom side of a F4B substrate, which has to fabricate). For maximization of the size reduction, the
thickness of 2 mm and dielectric constant of 2.65. A single vertical vias should be staggered arranged. Staggered vias at
spiral-shaped patch and a 50 Ω microstrip line are printed the cross-corners can make the distance of two adjacent vias
on the top side of the same substrate. This line is coaxial larger than center arranged vias. Thus, staggered arranged
fed by a 50 Ω SMA (SubMiniature version A) connector vias have smaller coupling capacitances than center arranged
underneath the ground plane. The feeding probe has radius vias; then the effect of cancellation to the inductances is
of 0.5 mm. In order to reduce the lateral size of this antenna, smaller than center arranged vias too. Therefore, for the
a novel technique is applied. This technique is implemented whole antenna with the same length, the staggered vias at
by separating the conventional spiral patch in Figure 1(a) the cross-corners can have larger inductances than center
into a different number of segments (denoted by N). Then, arranged vias, and larger inductance leads to larger size
half of these segments are moved to the bottom side of reduction. The proposed antenna structure is depicted in
Figures 1(b) and 1(c). It is identical to the conventional spiral 0

antenna in Figure 1(a) when N is equal to 1. Each via has −5
the same radius of 0.3 mm. For the proposed antenna, the
physical length of the current flowing path is increased with −10
the number of segments. Therefore, its resonant frequency −15 N = 64

is decreased with the value N. As a result, a larger N can
−20 N = 128
achieve a smaller antenna size. To prove the effectiveness N = 32
of the proposed size reduction technique, five printed spiral −25 N = 16 N = 1
antennas with N equals to 1, 16, 32, 64, and 128 are fabricated −30
and tested. The other parameters of the antennas are: r1 = 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
15 mm, r2 = 25 mm, w = 5 mm, L1 = 7 mm, L2 = 86 mm, Figure 2: Simulated input return loss of these antennas.
a = 70 mm, b = 92 mm.
3. Results and Analysis

In this paper, we make a fully simulation through a MoM-
(Method of Moments-) based software, IE3D. Figure 2 shows
the simulated input return loss of these antennas. It is clearly
seen that the antenna is resonant at 1.49 GHz when N =
1. If N is increased to 16, 32, 64, and 128, the resonance
frequency is reduced to 1.31 GHz, 1.14 GHz, 0.89 GHz,
and 0.61 GHz, respectively. Therefore, the suppression in
resonant frequency is 12.1%, 23.5%, 40.3%, and 59.1%,
respectively. The reduction in patch size is 22.7%, 41.5%,
64.3%, and 85.5%. The impedance bandwidths (S11 ≤
−10 dB) of these five antennas are 12.3%, 11.6%, 8.7%, Figure 3: Photograph of these five antennas.
5.08%, and 2.3%, respectively.
Figure 3 shows the photograph of these five antennas; 0
they have the same structures as the simulated structures. −5
The antenna’s performance is measured by the E5071C
−10
Network Analyzer and the Near-Field Antenna Measurement
System, Satimo. Figure 4 shows the measured input return −15
loss of these antennas. It can be seen that the antenna is −20 N = 128
resonant at 1.53 GHz when N = 1. If N is increased to 16, 32, N = 64
−25
64, and 128, the resonance frequency is reduced to 1.37 GHz,
1.2 GHz, 0.92 GHz, and 0.61 GHz, respectively. Therefore, −30 N = 32
the reduction in resonant frequency is 10%, 22%, 40%, −35 N = 16 N =1
and 60%, respectively. The reduction in patch size is 19%,
−40
39%, 64%, and 84%. The impedance bandwidths (S11 ≤ 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8
−10 dB) of these five antennas are 14%, 12%, 9%, 5.1%, and ×109
2.6%, respectively. In summary, the percentage size reduction
for the proposed antenna is increased with the number of Figure 4: Measured input return loss of these antennas.
segments used. More reduction in size is achieved at the
cost of narrower bandwidth. This is expected because the
bandwidth of an antenna is related to the electrical volume
of its radiation. expected because the radiation efficiency of an antenna
Figure 5(a) to Figure 5(d) show the measured radiation is dropped with its electrical radiation area [10]. We can
patterns of the proposed antennas with N = 1 to 64 in plane see from Figure 5(b) to Figure 5(d) that copolarization and
φ = 0◦ and φ = 90◦ . Because of the small ground plane, cross-polarization components are very close to each other.
the radiation patterns for the antennas are bidirectional. The reason is that the increasing number of vias leads to
Patterns for the antenna with N = 128 are not provided the inconsistencies in the structure borders and makes more
since the lowest measurement frequency of our Near-Field cross-polarization components which cannot be canceled
Antenna Measurement System (0.8 GHz) is higher than the out. This will be solved in future research.
resonant frequency (0.61 GHz) of this antenna. The gain
at the resonance frequency 1.53 GHz is 1.2 dBi. When N is 4. Conclusion
increased to 16, 32, and 64, the gain is reduced to 0.4 dBi,
−0.45 dBi, and −1.5 dBi. Therefore, increasing the number A novel vertical meandering technique to reduce the lateral
of segments will decrease the antenna gain. This is also size of a printed spiral antenna is proposed. Five printed
X-pol φ = 90
0 0
10 10
330 30 330 30
0 0
X-pol φ = 0 X-pol φ = 0
−10 −10
300 −20 60 300 −20 60
−30 −30 X-pol φ = 90
−40 −40
Co-pol φ = 90
270 90 270 90
Co-pol φ = 0 Co-pol φ = 90
Co-pol φ = 0
120 240 120
240
210 150
210 150
180 180
(a) (b)
0
0 10
10 330 30
330 30 0
0
−10
Co-pol φ = 90 −10 X-pol φ = 90
300 −20 60
300 −20 60
−30
−30
X-pol φ = 0
−40
−40
270 90
270 90
X-pol φ = 0
Co-pol φ = 0 X-pol φ = 90
240 240 120

120
Co-pol φ = 0
Co-pol φ = 90
210 150 210 150

180 180
(c) (d)
Figure 5: (a) Measured radiation patterns with N = 1. (b) Measured radiation patterns with N = 16. (c) Measured radiation patterns with
N = 32. (d) Measured radiation patterns with N = 64.
spiral antennas with a different number of segments are Research and Development Program of China (863 Program,
constructed, tested, and analyzed. According to the results no. 2006AA04A106).
achieved, it is evident that the size of the antenna can be
significantly reduced by increasing its number of segments.
References
Acknowledgments [1] J. Rashed and C. T. Tai, “A new class of resonant antennas,”
IEEE Transactions on Antennas and Propagation, vol. 39, no. 9,
The authors wish to thank Dr. Ji Li and Weijun Hong for pp. 1428–1430, 1991.
their helpful comments on the work presented in this paper. [2] J. M. Kim and J. G. Yook, “Compact mender-type slot
This work was supported in part by Shenzhen Science and antennas,” in Proceedings of the IEEE Antennas and Propagation
Technology Planning Project for the Establishment of Key Society International Symposium, vol. 2, pp. 724–727, Boston,
Laboratory in 2009 (CXB 200903090021A) and Hi-Tech Mass, USA, July 2001.
[3] S. Dey and R. Mittra, “Compact microstrip patch antenna,”

Microwave and Optical Technology Letters, vol. 13, no. 1, pp.
12–14, 1996.
[4] K. L. Wong, C. L. Tang, and H. T. Chen, “A compact
meandered circular microstrip antenna with a shorting pin,”
Microwave and Optical Technology Letters, vol. 15, no. 3, pp.
147–149, 1997.
[5] C. K. Wu, K. L. Wong, and W. S. Chen, “Slot-coupled
meandered microstrip antenna for compact dual-frequency
operation,” Electronics Letters, vol. 34, no. 11, pp. 1047–1048,
1998.
[6] J. H. Lu and K. L. Wong, “Slot-loaded, meandered rectangular
microstrip antenna with compact dual-frequency operation,”
[7] J. George, M. Deepukumar, C. K. Aanandan, P. Mohanan,
and K. G. Nair, “New compact microstrip antenna,” Electronics
Letters, vol. 32, no. 6, pp. 508–509, 1996.
[8] K. Noguchi, N. Yasui, M. Mizusawa, S. I. Betsudan, and T.
Katagi, “Increasing the bandwidth of a two-strip meander-
line antenna mounted on a conducting box,” in Proceedings
of the IEEE Antennas and Propagation Society International
Symposium, pp. 112–115, July 2001.
[9] H. Chaloupka, N. Klein, M. Peiniger, H. Piel, A. Pischke, and
G. Splitt, “Miniaturized high-temperature superconductor
microstrip patch antenna,” IEEE Transactions on Microwave
Theory and Techniques, vol. 39, pp. 1513–1521, 1991.
[10] L. J. Chu, “Physical limitations of omni-directional antennas,”
Journal of Applied Physics, vol. 19, no. 12, pp. 1163–1175, 1948.
doi:10.1155/2012/843754
Research Article
Microstrip Patch Antenna Bandwidth Enhancement Using
AMC/EBG Structures
R. C. Hadarig, M. E. de Cos, and F. Las-Heras

Area de Teorı́a de la Señal y Comunicaciones, Departamento de Ingenierı́a Eléctrica, Universidad de Oviedo, Edificio Polivalente,
Modulo 8, Campus Universitario de Gijón, Asturias, 33203 Gijón, Spain
Correspondence should be addressed to M. E. de Cos, medecos@tsc.uniovi.es
Received 11 July 2011; Revised 19 September 2011; Accepted 22 September 2011
Academic Editor: Zhongxiang Q. Shen
Copyright © 2012 R. C. Hadarig et al. This is an open access article distributed under the Creative Commons Attribution License,
A microstrip patch antenna with bandwidth enhancement by means of artificial magnetic conductor (AMC)/electromagnetic
band-gap structure (EGB) is presented. The electrical characteristics of the embedded structure are evaluated using MoM
simulations. The manufactured prototypes are characterized in terms of return loss, gain, and radiation pattern measurements
in an anechoic chamber.
1. Introduction The surface waves can be minimized using electromagnetic

band-gap structures whereas for obtaining a high gain anten-
Microstrip patch antennas offer an attractive solution to na an array configuration for the patch elements is needed.
compact and ease-low-cost design of modern wireless com- The research in the field of electromagnetic band-gap
munication systems due to their many advantages as light structures has become attractive in the antenna community
weight and low volume, low profile, planar configuration and is considered to be a key technology for improving mi-
which can be easily made conformal to host surface, low fab- crostrip patch antenna performances [4–6]. The use of meta-
rication cost, and the capability of obtaining dual and triple materials, such as the frequency selective surfaces (FSS) [7–
frequency operations. When mounted on rigid surfaces mi- 9] is an alternative to face antennas and microwave circuit
crostrip patch antennas are mechanically robust and can be problems and can provide either EBG or AMC behavior.
easily integrated with microwave integrated circuits (MICs). In previous works, several narrow band antennas have been
However, microstrip patch antennas suffer from a num- mounted on EBG/AMC structures [10–18].
ber of disadvantages as compared to conventional nonprint- Depending on the intended application, the 2.4 GHz
ed antennas. Some of their major drawbacks are the narrow frequency band can be used, for example, for wireless com-
bandwidth, low gain, and surface wave excitation that re- munications at 2.45 GHz or for RFID systems at 2.48 GHz.
duce radiation efficiency. To overcome one of their more In this paper, the main goal is to improve the bandwidth and
critical restrictions, narrow bandwidth, several techniques the radiation properties of a microstrip patch antenna in the
can be used [1]. First of all, a thicker substrate with a low die- 2.48 GHz band using two different approaches: combination
lectric constant or a ferrite composition provides a wider of the patch antenna with an EBG structure in the same
bandwidth but the first approach leads to no low-profile layer and combination of the patch antenna with an AMC
designs and increased in size, whereas the second solution is structure in two different layers. The aim of this work is
expensive. Secondly, noncontacting feeding methods such as challenging because two resonant structures are involved and
proximity/aperture coupled can be used to improve the im- when integrated together their resonant behavior is mutually
pedance bandwidth, but this is difficult to fabricate. Another influenced. Firstly, the design of a microstrip patch antenna,
possibility is multiresonator stack configuration with the henceforth referred as patch antenna, is shown followed
inconvenient of resulting large thickness prototype [2, 3]. by an adaptation of an AMC design recently presented by
W p = 65.52 mm
35.44 mm
L p = 81.9 mm
26.15 mm
(a) (b) (c)
Figure 1: Manufactured prototypes: (a) Patch antenna, (b) Patch antenna-EBG, (c) Planar AMC.
the authors to operate at 2.48 GHz. Then, the patch anten- is represented in Figure 2. The AMC resonant frequency is
na is placed above the AMC. This combination will be hence- 2.48 GHz, and the AMC operation bandwidth is approxi-
forth referred as Patch antenna-AMC. Secondly, the AMC mately 130 MHz (5.24%) (see Figure 2(b)). The structure ex-
structure is modified to act as an EBG at a frequency close hibits several advantages such as uniplanar feature since nei-
to the patch antenna resonance frequency. Finally the EBG is ther multilayer substrate no via holes are required, simpli-
combined with the patch antenna on the same layer, resulting fying the implementation and reducing its costs.
in a design with a uniplanar feature and reduced cost. This
combination will be henceforth referred as patch antenna- 2.2. EBG Characterization. The periodic structure can be
EBG. Return loss, gain, and radiation patterns of the three characterized as EBG using the suspended strip method [21,
prototypes (all having the same dimensions) are analyzed 22] (See Figure 3). A suspended strip line over the 4 × 4 cell
based on measurements in an anechoic chamber. arrangement is used to test the transmission response of the
electromagnetic waves. The strip height is 0.02 λ. The struc-
2. Microstrip Antenna Design ture will block the transmission of power along the strip line
for frequencies within the band gap region and a notice-
The microstrip patch antenna is a narrow band design. In
able reduction in S21 can be observed at a certain frequency
this work, the patch antenna suitable for RFID applications
band. The band-gap of the EBG lattice is designed to
at 2.48 GHz is designed using ROGER3010 substrate with
be adjacent to the frequency band of the patch antenna,
a thickness of 1.27 mm, relative dielectric permittivity εr =
so that when integrating the two structures on the same
10.2, and loss tangent of 0.0023.The geometry of the patch
layer, their resonances couple each other, and, as a result, a
antenna with its dimensions is shown in Figure 1(a). The
wider bandwidth will be generated without disturbing other
characteristic impedance of the transmission line is 50 Ω.
characteristics of the patch antenna such as the radiation
The antenna design has been carried out by a set of method-
pattern. The dimension of the unit cell in the case of EBG
of-moments (MoM) simulations with commercial software
characterization is W2 = 16.38 mm. The simulated band gap
[19]. From Figure 4, it can be extracted that the simulated
of the EBG structure closer to the patch antenna bandwidth
operating bandwidth of the patch antenna is 20 MHz.
is 45 MHz around 2.5 GHz (see Figure 4).
2.1. AMC Characterization. An adaptation of the AMC
previously designed by the authors [20] is carried out shifting 3. Microstrip Patch Antenna Combined
the resonant frequency to 2.48 GHz. Based on the Bloch- with AMC/EBG
Floquet theory and on the finite element method (FEM), a
single cell of the lattice with periodic boundary conditions 3.1. Patch Antenna Placed above the AMC Structure. A 4 ×
(PBCs) on its four sides is simulated in order to obtain the 5 cells planar AMC structure is placed as patch antenna
frequency band where the periodic structure acts as an AMC. ground plane [23] (see Figure 5) in order to analyze if the
The phase of the reflection coefficient on the AMC sur- antenna’s bandwidth and the radiation properties can be im-
face is computed using a uniform incident plane wave (see proved. The antenna is fixed to the AMC structure (see
Figure 2(a)). Depending on the unit cell geometry together Figure 1(c)) by a 0.1 mm double-sided nonconducting ad-
with the substrate’s thickness and relative dielectric permit- hesive tape. The microstrip patch antenna bandwidth is
tivity, the resonant frequency and the bandwidth of the struc- 20 MHz whereas the AMC operation bandwidth is 130 MHz,
ture can be tuned. The unit cell dimensions are W × W = having each one the same resonance frequency, 2.48 GHz,
16.93 × 16.93 mm2 and its geometry exhibits four sym- and the same dimensions. However, for combining the two
metry planes. The simulated reflection phase of normally structures, the antenna’s ground plane has been removed and
incident plane wave on the AMC surface versus frequency is placed above the AMC. As a consequence the antenna’s
Wave port
PBC 180 AMC BW

135
PBC
Reflection phase (deg)

90
45
W 0
W h −45
−90
−135 2.48 GHz SHF band

−180
RO3010
Unit cell
Conductor 2 2.2 2.4 2.6 2.8 3
ground plane Frequency (GHz)
(a) (b)
Figure 2: AMC unit cell: (a) reflection phase simulation setup, (b) phase of the reflection coefficient on the AMC surface.
Metal 0
(cooper)
−5
−10
|Si j | (dB)
−15
−20
−25
−30
2.4 2.44 2.48 2.52 2.56 2.6 2.64
Suspended Frequency (GHz)
line
S11 patch antenna (MoM simulation)
S21 EBG-suspended line structure (MoM simulation)
Figure 3: Schematic of suspended line above EBG surface (top
view). Figure 4: Resonances to be coupled in order to achieve bandwidth
enhancement.
resonance frequency decreases due to capacitive effects for 3.2. Patch Antenna Surrounded by the EBG Structure. In
those frequencies within the AMC bandwidth. A resonance order to suppress the surface waves and to increase the band-
is obtained in the AMC bandwidth and outside this band width by means of coupled resonators effect, the EBG lattice
the antenna behaves as if its substrate thickness had doubled. is arranged around the patch, forming a uniplanar design
Merging both effects, the combined structure resonates in a [24]. As it has been already mentioned in Section 2.2, the
bandwidth wider than the microstrip patch antenna alone, resonance frequency of both structures (patch antenna and
but narrower than the AMC bandwidth. As disadvantage, EBG structure) is mutually influenced, and depending on the
the thickness of the combined structure is increased. If the frequency difference between them and the unit-cell arrange-
dielectric substrate thickness of the patch antenna doubles, ment around the patch antenna, the resulting resonance fre-
the resulting bandwidth (30 MHz) is narrower compared to quency changes. The frequencies included on the patch
the bandwidth of patch antenna-EBG and patch antenna- antenna’s bandwidth are adjacent to the ones included on
AMC prototypes. the lower band gap. The selected EBG arrangement with
−5
−10
−15
|S11 | (dB)
ROGER3010 −20
−25
−30
AMC BW
−35
2.35 2.4 2.42 2.45 2.5 2.55
Frequency (GHz)
Patch antenna
Patch antenna-EBG
ROGER3010
Patch antenna-AMC
Figure 5: Patch antenna-AMC prototype layout. Figure 7: Measurement comparison between the prototypes: patch
antenna, patch antenna-EBG, and patch antenna-AMC.
0
−5 for the dielectric substrate, or even more likely due to manu-
−10 facturing tolerances.
In the case of placing the antenna above the AMC surface
−15 the antenna resonance frequency is shifted downwards to
|S11 | (dB)
−20 2.43 GHz (see Figure 7) due to the capacitive effects that are
generated between the two combined structures. Also, as the
−25
AMC structure has wider bandwidth than the patch anten-
−30 na, the resulting prototype bandwidth increases to 46 MHz,
−35 meaning a 100% broader bandwidth (see Figure 7).
When the patch antenna is surrounded by one row of
−40
2.4 2.42 2.44 2.46 2.48 2.5 2.52 2.54 EBG cells the bandwidth increases 50% (see Figure 7) due
Frequency (GHz) to the property of coupling the frequency bands of the two
structures composing the prototype. It is remarkable that
Patch antenna (measurement) this 50% bandwidth improvement is achieved increasing
Patch antenna (MoM simulation) neither the prototype size nor the thickness. The percentage
Patch antenna-EBG (measurement)
bandwidth comparison of the three prototypes is presented
Patch antenna-EBG (MoM simulation)
in Table 1.
Figure 6: Simulation and measurement comparison between the Measured radiation pattern cuts in the E and H planes
prototypes: patch antenna and patch antenna-EBG. of each manufactured prototype are presented in Figure 8.
The patch antenna prototype exhibits copolarization-cross-
polarization (CP-XP) ratio better than 25 dB (see Table 2),
whereas for the patch antenna-EBG prototype the (CP-XP)
respect to the antenna is a tradeoff between performance and ratio and the directivity are even increased. In measurements
size. The dimensions of the final structure (Figure 1(b)) are the gain of the patch antenna (4.59 dB) is preserved when
W p = 65.52 mm and L p = 81.90 mm. the antenna is surrounded by one row of EBG cells (Table 1).
From the simulation results, using EGB structures around
4. Results the patch antenna its radiation efficiency increases, due to
surface wave suppression property. However from measure-
Prototypes of the patch antenna, patch antenna placed above ment results it can be concluded that for this specific ar-
the AMC surface, and patch antenna surrounded by the EBG rangement, the radiation efficiency is preserved (while im-
cells have been manufactured using laser micromachining. proving bandwidth). The difference between simulations and
The return losses of each manufactured prototype have been measurements relies on the fact that the simulation method
measured. As it can be observed in Figure 6 the measured implemented by Momentum considers infinite dielectric
operating bandwidth of the patch antenna is 23 MHz. The under the finite EBG metallization but also the difference
difference in bandwidth between simulations (20 MHz) and could be attributable to misalignments in the anechoic
measurement (23 MHz) results could be due to the fact that chamber. Radiation pattern properties of the patch antenna-
the commercial MoM software considers infinite extension AMC prototype show a (CP-XP) ratio inferior to the other
E plane. Normalized amplitude (dB) H plane. Normalized amplitude (dB)

90 90 0 dB
0 dB
120 60 120 60
−10 dB
−10 dB
150 30 150 30
−20 dB
−20 dB
−30 dB
−30 dB
180 0 180 0
210 330 210 330
240 300 240 300

270 270
Patch antenna, CP f = 2.48 GHz Patch antenna, CP f = 2.48 GHz
Patch antenna, XP f = 2.48 GHz Patch antenna, XP f = 2.48 GHz
Patch antenna-EBG, CP f = 2.49 GHz Patch antenna-EBG, CP f = 2.49 GHz
Patch antenna-EBG, XP f = 2.49 GHz Patch antenna-EBG, XP f = 2.49 GHz
Patch antenna-AMC, CP f = 2.43 GHz Patch antenna-AMC, CP f = 2.43 GHz
Patch antenna-AMC, XP f = 2.43 GHz Patch antenna-AMC, XP f = 2.43 GHz
(a) (b)
Figure 8: Measured radiation patterns of the prototypes: Patch antenna, patch antenna-EBG, and patch antenna-AMC.
Table 1: Comparison between the three designs.
Prototype Bandwidth (MHz) Directivity (dB) Gain (dB) Radiation efficiency (%)
Meas. Meas. Sim. Meas. Sim. Meas. Sim.
Patch antenna 23 (0.93%) 7.33 5.95 4.59 4.29 53.21 68.23
Patch antenna-EBG 34 (1.37%) 7.50 6.84 4.61 5.56 51.40 74.47
Patch antenna-AMC 46 (1.90%) 6.72 8.52 0 0.79 21.28 16.86
two prototypes and a gain close to 0 dB. As the AMC does not Table 2: CP-XP ratio comparison.
have the ability to suppress the surface waves and the fact that
CP-XP ratio CP-XP ratio
the thicker the substrate, the stronger are the surface waves, Prototype
(E plane, dB) (H plane, dB)
the gain of the patch antenna-AMC prototype does not
improve. Also as the CP-XP ratio is worst for patch antenna- Patch antenna 25.81 25.04
AMC than for the other prototypes, part of the energy could Patch antenna-EBG 30.43 28.79
be radiated in other polarizations and backwards (as it is Patch antenna-AMC 13.85 9.43
shown in Figure 8). In order to improve the gain, a gap be-
tween the antenna and the AMC surface could be used but
this is technologically less advantageous. In addition, the
microstrip patch antenna’s gain and directivity can be in- microstrip patch antenna in order to characterize their joint
creased when more rows or/and columns surround the performance. The prototypes have been manufactured and
prototype, so a trade-off between performance and size must characterized based on measurements in an anechoic cham-
be taken (the higher the number of unit cells in a periodic ar- ber. From the measurements of the two resulting prototypes,
rangement, the closer its behavior to an infinite EBG struc- patch antenna-AMC and patch antenna-EBG, it can be
ture). concluded that both prototypes improve the bandwidth of
the patch antenna. Due to the surface wave effect of the EBG
5. Conclusions structure, the patch antenna-EBG prototype shows better
radiation properties increasing neither the prototype size
Bandwidth enhancement of microstrip patch antenna in the nor the thickness. All the prototypes presented are robust,
RFID SHF 2.48 GHz band has been presented. Two different compact and do not require via holes, being compatible with
structures (AMC/EBG) have been combined with the same planar fabrication technology.
Acknowledgments IEEE Transactions on Antennas and Propagation, vol. 55, no.

6, pp. 1691–1697, 2007.
This work has been supported by the Ministerio de Ciencia [15] G. Poilasne, “Antennas on High impedance ground planes: on
e Innovación of Spain/FEDER under Projects TEC2008- the importance of the antenna isolation,” Progress in Electro-
01638/TEC (INVEMTA), and CONSOLIDER-INGENIO magnetics Research, vol. 41, pp. 237–255, 2003.
CSD2008-00068 (TERASENSE), by the Gobierno del Prin- [16] S. Zhu and R. Langley, “Dual-band wearable textile antenna
cipado de Asturias (PCTI)/FEDER-FSE under Projects on an EBG substrate,” IEEE Transactions on Antennas and
EQUIP08-06, FC09-COF09-12, EQUIP10-31, and PC10-06 Propagation, vol. 57, no. 4, pp. 926–935, 2009.
(FLEXANT), by Grant BP10-039, and by Cátedra Telefónica, [17] M. Mantash, A. C. Tarot, S. Collardey, and K. Mahdjoubi,
“Dual-band antenna for WLAN application with EBG,” in
Universidad de Oviedo.
the 4th International Congress on Advanced Electromagnetic
Materials in Microwaves and Optics, pp. 794–796, Karlsruhe,
References Germany, September 2010.
[18] M. Mantash, A. C. Tarot, S. Collardey, and K. Mahdjoubi,
[1] R. Garg, I. Bhartia, I. Bahl, and A. Ittipiboon, Microstrip An- “Dual-band CPW-fed G-antenna using an EBG structure,” in
tenna Design Handbook, Artech House, Boston, Mass, USA, Antennas and Propagation Conference (LAPC), pp. 453–456,
2001. Loughborough, UK, 2010.
[2] G. Kumar and K. C. Gupta, “Directly coupled multiple reso- [19] “ADS Momentum EM simulator tool,” http://www.home
nator wide-band microstrip antenna,” IEEE Transactions on .agilent.com/agilent/product.jspx?cc=EG&lc=eng&ckey=
Antennas and Propagation, vol. 33, no. 6, pp. 588–593, 1985. 1475688&nid=-34360.0.00&id=1475688.
[3] F. Yang, X. X. Zhang, X. Ye, and Y. Rahmat-Samii, “Wide-band [20] M. E. De Cos, Y. Álvarez, R. C. Hadarig, and F. Las-Heras,
E-shaped patch antennas for wireless communications,” IEEE “Novel SHF-band uniplanar artificial magnetic conductor,”
Transactions on Antennas and Propagation, vol. 49, no. 7, pp. IEEE Antennas and Wireless Propagation Letters, vol. 9, pp. 44–
1094–1100, 2001. 47, 2010.
[4] A. Pirhadi, F. Keshmiri, M. Hakkak, and M. Tayarani, “Analysis [21] L. Yang, M. Fan, F. Chen, J. She, and Z. Feng, “A novel com-
and design of dual band high directivity EBG resonator anten- pact electromagnetic-bandgap (EBG) structure and its ap-
na using square loop FSS AS superstrate layer,” Progress in plications for microwave circuits,” IEEE Transactions on Micro-
Electromagnetics Research, vol. 70, pp. 1–20, 2007. wave Theory and Techniques, vol. 53, no. 1, pp. 183–189, 2005.
[5] E. Rajo-Iglesias, L. Inclán-Sánchez, and O. Quevedo-Teruel, [22] A. Aminian, F. Yang, and Y. Rahmat-Samii, “In-phase reflec-
“Back radiation reduction in patch antennas using planar soft tion and EM wave suppression characteristics of electromag-
surfaces,” Progress In Electromagnetics Research Letters, vol. 6, netic band gap ground planes,” in IEEE International Antennas
pp. 123–130, 2009. and Propagation Symposium and USNC/CNC/URSI North
[6] Z. Duan, S. Qu, and Y. Hou, “Electrically small antenna in- American Radio Science Meeting, pp. 430–433, June 2003.
spired by spired split ring resonator,” Progress In Electromag- [23] R. C. Hadarig, M. E. De Cos, Y. ÁLvarez, and F. Las-Heras,
netics Research Letters, vol. 7, pp. 47–57, 2009. “Novel bow-tie antenna on artificial magnetic conductor
[7] F. Yang and Y. Rahmat-Samii, Electromagnetic Band-Gap for 5.8 GHz radio frequency identification tags usable with
Structures in Antenna Engineering, The Cambridge RF and metallic objects,” IET Microwaves, Antennas and Propagation,
Microwave Engineering Series, Cambridge University Press, vol. 5, no. 9, pp. 1097–1102, 2011.
Cambridge, Mass, USA, 2008. [24] M. E. De Cos, Y. Álvarez, and F. Las-Heras, “Enhancing
[8] M. E. De Cos, F. L. Heras, and M. Franco, “Design of planar patch antenna bandwidth by means of uniplanar EBG-AMC,”
artificial magnetic conductor ground plane using frequency- Microwave and Optical Technology Letters, vol. 53, no. 6, pp.
selective surfaces for frequencies below 1 GHz,” IEEE Antennas 1372–1377, 2011.
and Wireless Propagation Letters, vol. 8, pp. 951–954, 2009.
[9] O. Luukkonen, C. R. Simovski, and S. A. Tretyakov, “Ground-
ed uniaxial material slabs as magnetic conductors,” Progress in
Electromagnetics Research B, no. 15, pp. 267–283, 2009.
[10] H. Shaban, H. Elmikaty, and A. A. Shaalan, “Study the effects
of electromagnetic band-gap (EBG) substrate on two patch
microstrip antenna,” Progress in Electromagnetics Research B,
vol. 10, pp. 55–74, 2008.
[11] F. Yang and Y. Rahmat-Samii, “Reflection phase characteriza-
tions of the EBG ground plane for low profile wire antenna,”
IEEE Transactions on Antennas and Propagation, vol. 51, no.
10, pp. 2691–2703, 2003.
[12] J. R. Sohn, K. Y. Kim, H. S. Tae, and J. H. Lee, “Comparative
study on various artificial magnetic conductors for low-profile
antenna,” Progress in Electromagnetics Research, vol. 61, pp. 27–
37, 2006.
[13] S. Chaimool, K. L. Chung, and P. Akkaraekthalin, “Bandwidth
and gain enhancement of microstrip patch antennas using
reflective metasurface,” IEICE Transactions on Communica-
tions, vol. E93-B, no. 10, pp. 2496–2503, 2010.
[14] J. Liang and H. Y. D. Yang, “Radiation characteristics of a
microstrip patch over an electromagnetic bandgap surface,”
doi:10.1155/2012/435890
Research Article
High-Performance Computational Electromagnetic Methods
Applied to the Design of Patch Antenna with EBG Structure
R. C. Hadarig, M. E. de Cos, and F. Las-Heras

Area de Teorı́a de la Señal y Comunicaciones, Departamento de Ingenierı́a Eléctrica, Universidad de Oviedo, Edificio Polivalente,
Modulo 8, Campus Universitario de Gijón, 33203 Gijón, Spain
Correspondence should be addressed to M. E. de Cos, medecos@tsc.uniovi.es
Received 14 June 2011; Revised 19 September 2011; Accepted 20 September 2011
Academic Editor: Shyh-Kang Jeng
Copyright © 2012 R. C. Hadarig et al. This is an open access article distributed under the Creative Commons Attribution License,
In this contribution High-Performance Computing electromagnetic methods are applied to the design of a patch antenna
combined with EBG structure in order to obtain bandwidth enhancement. The electrical characteristics of the embedded structure
(patch antenna surrounded by EBG unit cells) are evaluated by means of method of moment technique (MoM) whereas for
designing the unit cell, the finite element method (FEM) together with the Bloch-Floquet theory is used. The manufactured
prototypes are characterized in terms of return loss and radiation pattern in an anechoic chamber.
1. Introduction mental roll in order to get a proper design in a reasonable

time.
Microstrip patch antennas, which are used for both defense
and commercial applications, are replacing many conven-
tional notprinted antennas. The most important properties 2. Microstrip Antenna
of microstrip patch antennas are the lightweight, small- The antenna design has been carried out using Method of
volume and the mass production at low cost whereas Moments (Momentum) electromagnetic simulator [17] and
intrinsic disadvantages that limit their applications are low its geometry (Figure 1(a)) was optimized varying the values
gain, narrow bandwidth, and excitation of surface waves in a continuous range in order to obtain the frequency
[1]. In order to design antennas with better efficiency and band of interest (2.48 GHz) and the minimum bandwidth
gain and lower backlobe and sidelobe levels, EBG structures requested (the process is an iterative one in which there will
can be used [2–7]. In previous works [8–16] several narrow be a tradeoff between the variation of antenna’s geometry
band antennas have been mounted on EBG structures. In and its effects on the performance). Using parameter sweep
this contribution, for obtaining bandwidth enhancement with MoM in a 2 core Intel Xeon X5560 48 GB RAM
in the 2.48 GHz band, the band-gap of the EBG lattice is (equivalent to 16 execution threads) server, there can be
designed to be adjacent to the frequency band of the patch obtained the solution for 81 different values of the swept
antenna, so that when integrating together both structures parameter in just one minute. This is due to a mesh definition
their resonances couple each other and as a result a wider so that 20 cells per wavelength at 3 GHz are taken which
bandwidth will be generated. The final design is uniplanar leads to 249 rectangular cells and 171 triangular cells with
and in addition there is no need for via holes. a matrix size of 740. After applying mesh reduction a matrix
The aim of this work is challenging because two resonant size of 610 results. A dielectric substrate, ROGER3010 having
structures are involved and when integrated together their 1.27 mm thickness, εr = 10.2 relative dielectric permittivity,
resonant behavior is mutually influenced. For this reason and loss tangent of 0.0023, has been used. The measured
the precision in the simulation results is a key point to operating bandwidth of the patch antenna is 23 MHz
achieve bandwidth enhancement. At this point is where high- (Figure 4). The difference in bandwidth between the sim-
performance computational electromagnetics play a funda- ulated (20 MHz) and measured (23 MHz) results could be
65.52 mm Conductor w
ground plane h
w
Metal
(cooper)
35.44 mm
26.15 mm Unit
81.9 mm
cell
(a) (b)
Figure 1: Manufactured prototypes: (a) patch antenna (b) patch

antenna-EBG. Suspended
line
due to the fact that the commercial MoM software considers

Figure 2: Schematic of suspended line above EBG surface (top
infinite extension for the dielectric substrate, or even more
view).
likely due to manufacturing tolerances.
3. EBG Characterization that needs to be simulated is reduced), for the current

scenario the method of moments (MoM) could yield a
The unit cell of an EBG lattice consists of metal pads and faster computational time (using MoM only the conductive
sometimes narrow lines that implement a distributed LC structure is meshed whereas using FEM, the whole volume
circuit having a resonant frequency [18]. In [19] the same around the structure including the air box is meshed). The
unit cell but in different frequency band is characterized from periodic structure can be characterized as EBG using the
the point of view of an AMC, computing the phase of the suspended strip method [20, 21] and so the transmission
reflection coefficient on its surface, from an uniform incident coefficient (S21 ) of a suspended strip line over the periodic
plane wave. To search for the frequency band in which structure is simulated (Figure 2). The structure will block
the periodic structure shows the AMC behavior, the finite the transmission of power along the strip line for frequencies
element method (FEM) together with the Bloch-Floquet within the band-gap region and a noticeable reduction of S21
theory is used. A single cell of the lattice (with periodic can be observed at a certain frequency band. To accomplish
boundary conditions (PBCs) on its four sides) is simulated this simulation the mesh has been defined so that 20 cells per
to model an infinite structure. The unit cell dimensions are wavelength at 3 GHz are taken leading to 2085 rectangular
W × W = 16.38 × 16.38 mm2 . The simulation is carried out cells and 5008 triangular cells with a matrix size of 9706. After
using a server with 2 processors Intel Xeon E5620 and 64 GB applying mesh reduction results a matrix size of 2352. The
RAM in a configuration equivalent to 2 execution threats and simulation time for 25 frequency steps is 50 minutes in a 2-
32 GB RAM. The air-box size is λ/2 at the lower frequency core Intel Xeon X5560 48 GB RAM (equivalent to 16 threads
considered in the simulation (which is 1 GHz). The solution execution) server.
frequency is 3 GHz and a frequency sweep is carried out from The band-gap of the EBG structure (45 MHz) is adjacent
1 GHz to 3 GHz with a 0.01 GHz frequency step. The mesh is to the bandwidth of the patch antenna (see Figure 3); thus
established to take at least 10 tetrahedra per wavelength at when combining the two structures together, bandwidth
the solution frequency (3 GHz). Mixed-order basis functions enhancement is obtained without disturbing other charac-
and 30% lambda refinement are used. Maximum Delta S is teristics of the patch antenna, such as the radiation pattern.
fixed to 0.02 for the S-parameter calculations. All this leads to
a mesh with 3642 tetrahedra (2889 for the air-box and 753 for 4. Patch Antenna with EBG Structure
the substrate), a matrix size of 16101, and a computational
time of 8 minutes and 44 seconds for the aforementioned Once the antenna and the EBG structures have been
frequency sweep. designed, the next step is the integration of both resonant
However, for the intended application, the structure structures together in the same layer forming a uniplanar
should show EBG behavior. Even though the finite element design [22]. As it has been already mentioned in the intro-
method (FEM) uses specific boundary conditions such as duction, the resonance frequency of both structures is mutu-
simulating just half the volume under study (using Perfect ally influenced, and so depending on the frequency difference
Magnetic Conductor (PMC) boundary condition in one of between them, and how the unit-cell arrangement around
the volume walls, the one that would cut the volume in the patch antenna is, the resulting resonance frequency will
two identical parts, the size of the electromagnetic problem change. The design of the patch antenna surrounded by
−5
−10
|Si j | (dB)
−15
−20
−25
−30
2.4 2.44 2.48 2.52 2.56 2.6 2.64
Frequency (GHz)
S11 patch antenna (MoM simulation)

S21 EBG-suspended line structure (MoM simulation)
Figure 3: Resonances to be coupled in order to achieve bandwidth enhancement.
−5
−10
|S11 | (dB)
−15
−20
−25
−30
−35
2.42 2.44 2.46 2.48 2.5 2.52 2.54 2.56
Frequency (GHz)
Patch antenna (measurement)

Patch antenna (MoM simulation)
Patch antenna-EBG (measurement)
Patch antenna-EBG (MoM simulation)
Figure 4: Simulation and measurement comparison between the patch antenna and the patch antenna-EBG prototypes.
one row EBG lattice has been carried out (Figure 1(b)). In frequency steps is 63 minutes in a 2-core Intel Xeon X5560
FEM, due to the air-box size when small frequencies are 48 GB RAM (equivalent to 16 threads execution) server.
involved (as in this case), the electric size of the problem to A prototype of the Patch antenna-EBG has been manu-
be solved is rather big. A proper mesh should take at least factured using laser micromachining. The return losses of the
10 (generally 20) tetrahedra per wavelength. Depending on manufactured prototype have been measured and compared
the prototype’s physical size, this could make the matrix of to those of the microstrip patch antenna (Figure 4). The
the linear equation system to become dense, which is not principle of achieving bandwidth enhancement is based
desirable in FEM and generally leads to longer computational on coupling the resonances of the structures involved. As
times and increased memory requirements. However the the patch antenna resonates at adjacent frequency band
matrix in MoM is dense, so this is not a problem, and thus compared to the band-gap of the EBG lattice, a signifi-
this could be a better choice in general. The disadvantage cant bandwidth enhancement of the prototype combining
of MoM is related to dielectric managements as they are the two structures (Patch antenna-EBG) is obtained. As
considered infinite sized. Using MoM the mesh has been shown in Figure 4 the resulting bandwidth (34 MHz) of
defined so that 20 cells per wavelength at 3 GHz are taken the Patch antenna-EBG is 50% wider than the microstrip
which leads to 2460 rectangular cells and 6246 triangular cells patch antenna’s bandwidth (23 MHz). Radiation pattern
with a matrix size of 12365. After applying mesh reduction measurements of the prototypes have been carried out
a matrix size of 5832 results. The simulation time for 81 in anechoic chamber to complete their characterization.
E plane. Normalized amplitude (dB) H plane. Normalized amplitude (dB)

90 0 90 0
120 60 120 60
−10 −10
150 −20 30 150 −20 30
−30 −30
180 0 180 0
210 330 210 330
240 300 240 300

270 270
Patch antenna, CP Patch antenna-EBG, CP Patch antenna, CP Patch antenna-EBG, CP

Patch antenna, XP Patch antenna-EBG, XP Patch antenna, XP Patch antenna-EBG, XP
(a) (b)
Figure 5: Patch antenna and patch antenna-EBG-measured radiation patterns.
Table 1: Comparison between the two prototypes. there are also other possible approaches, time domain based
such as (Finite-difference time-domain) FDTD which could
Prototype Bandwidth (MHz) Directivity (dB) Gain (dB)
also be used.
Patch antenna 23 7.33 4.59
There was reported a 50% increase of the initial band-
Patch antenna-EBG 34 7.50 4.61 width. The patch antenna-EBG prototype presented in
this paper has several advantages such as planar feature,
compact size, and low dielectric losses. Neither via holes
Radiation pattern cuts in the E and H planes of each nor multilayer substrates are required, simplifying practical
manufactured prototype are plotted in Figure 5. Using an implementation and reducing its cost.
EBG structure to surround the patch antenna, the directivity
increases due to the surface wave suppression (Table 1).
In the case of placing the patch antenna in a frequency Acknowledgments
range outside the band-gap of the EBG structure, MoM This work has been supported by the Ministerio de Ciencia
simulations show that the unit cells have no influence in the e Innovación of Spain/FEDER under projects TEC2008-
radiation properties or in the bandwidth. 01638/TEC (INVEMTA) and CONSOLIDER-INGENIO
CSD2008-00068 (TERASENSE), by the Gobierno del Prin-
5. Conclusions cipado de Asturias (PCTI)/FEDER-FSE under projects
EQUIP08-06, FC09-COF09-12, EQUIP10-31, and PC10-06
Bandwidth enhancement of microstrip patch antenna by (FLEXANT), by grant BP10-039, and by Cátedra Telefónica-
means of EBG structure for RFID SHF 2.48 GHz band Universidad de Oviedo.
has been presented. Using High-Performance computing
electromagnetic methods the electrical characteristics of the
resonant unit cell and the patch antenna have been evaluated References
both separated as well as combined in the same layer. The [1] R. Garg, I. Bhartia, I. Bahl, and A. Ittipiboon, Microstrip
simulated and measured results are in good agreement due Antenna Design Handbook, Artech House, Boston, Mass, USA,
to precision of the methods (MoM and FEM) used in 2001.
simulations. Both frequency domain techniques, MoM and [2] A. Pirhadi, F. Keshmiri, M. Hakkak, and M. Tayarani, “Analysis
FEM, can be used once the AMC is designed using FEM with and design of dual band high directivity EBG resonator
PBCs. This is just an example of MoM and FEM techniques antenna using square loop FSS AS superstrate layer,” Progress
application to the design of antennas with metamaterials, but in Electromagnetics Research, vol. 70, pp. 1–20, 2007.
[3] E. Rajo-Iglesias, L. Inclán-Sánchez, and O. Quevedo-Teruel, [19] M. E. de Cos, Y. Álvarez, R. C. Hadarig, and F. Las-Heras,
“Back radiation reduction in patch antennas using planar soft “Novel SHF-band uniplanar artificial magnetic conductor,”
surfaces,” Progress in Electromagnetics Research Letters, vol. 6, IEEE Antennas and Wireless Propagation Letters, vol. 9, pp. 44–
pp. 123–130, 2009. 47, 2010.
[4] Z. Duan, S. Qu, and Y. Hou, “Electrically small antenna [20] A. Aminian, F. Yang, and Y. Rahmat-Samii, “In-phase reflec-
inspired by spired split ring resonator,” Progress in Electromag- tion and EM wave suppression characteristics of electromag-
netics Research Letters, vol. 7, pp. 47–57, 2009. netic band gap ground planes,” in IEEE International Antennas
[5] F. Yang and Y. Rahmat-Samii, Electromagnetic Band-Gap and Propagation Symposium, vol. 4, pp. 430–433, June 2003.
Structures in Antenna Engineering, The Cambridge RF and [21] L. Yang, M. Fan, F. Chen, J. She, and Z. Feng, “A novel
Microwave Engineering Series, Cambridge, University Press, compact electromagnetic-bandgap (EBG) structure and its
2008. applications for microwave circuits,” IEEE Transactions on
[6] M. E. de Cos, F. Las-Heras, and M. Franco, “Design of planar Microwave Theory and Techniques, vol. 53, no. 1, pp. 183–190,
artificial magnetic conductor ground plane using frequency- 2005.
selective surfaces for frequencies below 1 GHz,” IEEE Antennas [22] M. E. de Cos, Y. Álvarez, and F. Las-Heras, “Enhancing
and Wireless Propagation Letters, vol. 8, pp. 951–954, 2009. patch antenna bandwidth by means of uniplanar EBG-AMC,”
[7] O. Luukkonen, C. R. Simovski, and S. A. Tretyakov, Microwave and Optical Technology Letters, vol. 53, no. 6, pp.
“Grounded uniaxial material slabs as magnetic conductors,” 1372–1377, 2011.
Progress in Electromagnetics Research B, no. 15, pp. 267–283,
2009.
[8] H. Shaban, H. Elmikaty, and A. Shaalan, “Study the effects
of electromagnetic band-gap (EBG) substrate on two patch
microstrip antenna,” Progress in Electromagnetics Research B,
vol. 10, pp. 55–74, 2008.
[9] A. P. Feresidis, G. Goussetis, S. Wang, and J. C. Vardaxoglou,
“Artificial magnetic conductor surfaces and their application
to low-profile high-gain planar antennas,” IEEE Transactions
on Antennas and Propagation, vol. 53, no. 1, pp. 209–215, 2005.
[10] H. Mosallaei and K. Sarabandi, “Antenna miniaturization
and bandwidth enhancement using a reactive impedance
substrate,” IEEE Transactions on Antennas and Propagation,
vol. 52, no. 9, pp. 2403–2414, 2004.
[11] L. Akhoondzadeh-Asl, D. J. Kern, P. S. Hall, and D. H. Werner,
“Wideband dipoles on electromagnetic bandgap ground
planes,” IEEE Transactions on Antennas and Propagation, vol.
55, no. 9, pp. 2426–2434, 2007.
[12] J. Liang and H. Y. D. Yang, “Radiation characteristics of a
microstrip patch over an electromagnetic bandgap surface,”
IEEE Transactions on Antennas and Propagation, vol. 55, no.
6, pp. 1691–1697, 2007.
[13] F. Yang and Y. Rahmat-Samii, “Reflection phase characteriza-
tions of the EBG ground plane for low profile wire antenna
applications,” IEEE Transactions on Antennas and Propagation,
vol. 51, no. 10, pp. 2691–2703, 2003.
[14] J. R. Sohn, K. Y. Kim, H. S. Tae, and J. H. Lee, “Comparative
study on various artificial magnetic conductors for low-profile
antenna,” Progress in Electromagnetics Research, vol. 61, pp. 27–
37, 2006.
[15] S. Chaimool, K. L. Chung, and P. Akkaraekthalin, “Bandwidth
and gain enhancement of microstrip patch antennas using
reflective metasurface,” IEICE Transactions on Communica-
tions, vol. E93-B, no. 10, pp. 2496–2503, 2010.
[16] D. Nashaat, H. A. Elsadek, E. A. Abdallah, M. F. Iskander,
and H. M. Elhenawy, “Ultrawide bandwidth 2 × 2 microstrip
patch array antenna using electromagnetic band-gap structure
(EBG),” IEEE Transactions on Antennas and Propagation, vol.
59, no. 5, pp. 1528–1534, 2011.
[17] “ADS Momentum EM simulator tool,” http://www.agilent
.com/find/eesof.
[18] D. Sievenpiper, L. Zhang, R. F. J. Broas, N. G. Alexöpolous, and
E. Yablonovitch, “High-impedance electromagnetic surfaces
with a forbidden frequency band,” IEEE Transactions on
2074, 1999.
doi:10.1155/2012/595290
Application Article
A Wideband High-Gain Dual-Polarized Slot Array Patch Antenna
for WiMAX Applications in 5.8 GHz
Amir Reza Dastkhosh1 and Hamid Reza Dalili Oskouei2

1 Sahand University of Technology, P.O. Box 51335-1996, Tabriz, Iran
2 University of Aeronautical Science & Technology (Shahid Sattari), P.O. Box 13846-63113, Tehran, Iran
Correspondence should be addressed to Amir Reza Dastkhosh, amir reza dastkhosh@yahoo.com
Received 9 April 2011; Revised 12 May 2011; Accepted 20 July 2011
Academic Editor: Dau-Chyrh Chang
Copyright © 2012 A. R. Dastkhosh and H. Dalili Oskouei. This is an open access article distributed under the Creative Commons
Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is
properly cited.
A low-cost, easy-to-fabricate, wideband and high-gain dual-polarized array antenna employing an innovative microstrip slot patch
antenna element is designed and fabricated. The design parameters of the antenna are optimized using commercial softwares
(Microwave Office and Zeland IE3D) to get the suitable S-parameters and radiation patterns. Finally, the simulation results are
compared to the experimental ones and a good agreement is demonstrated. The antenna has an approximately bandwidth of 14%
(5.15–5.9 GHz) which covers Worldwide Interoperability Microwave Access (WiMAX)/5.8. It also has the peak gain of 26 dBi for
both polarizations and high isolation between two ports over a wide bandwidth.
1. Introduction This antenna has high isolation between the two ports over
a wide bandwidth more than 14%. Furthermore, this high-
Recently, microstrip patch antennas are one of the most gain (25.5 dBi) array antenna has dual polarization with
commonly used antenna types due to many advantages such a minimum half-power beamwidth (HPBW) (vertical: 7◦ ;
as light weight, low fabrication costs, planar configuration, horizontal: 6◦ ). The impedance characteristics, radiation
and capability to integrate with microwave integrated cir- pattern, return loss, and isolation between two ports for the
cuits. Thus, the patch antennas are very suitable for vari- designed array are analyzed, simulated, and optimized using
ous applications such as wireless communication systems, Microwave Office and Zeland IE3D softwares. Also, S11 , S21 ,
cellular phones, satellite communication systems, and radar and radiation pattern are measured and compared to the
systems [1–6]. Due to their inherent features they are found simulated ones.
attractive for applications in broadband networks. WiMAX
is a standard-based wireless technology for broadband
networks providing high data rate communication by using 2. Configuration of Element Antenna
low-cost equipment. The access points in this network are
usually built with large physical spacing. Therefore, the Microstrip patch antennas can be excited by different types
high-gain antenna is necessary to execute the long distance of feeds. In order to achieve the desired performances of
transmission with a lower power. WiMAX has three allocated WiMAX antenna, an aperture coupled feed is used because of
frequency bands called low band (2.5 GHz to 2.8 GHz), its good characteristics such as wide operational bandwidth
middle band (3.2 GHz to 3.8 GHz), and high band (5.2 GHz and shielding of the radiation patches. Moreover, an aperture
to 5.8 GHz). In this work, the low-cost microstrip slot array coupled feed yields better gain and radiation pattern for
antenna (8 × 8) is designed, simulated, and fabricated for a dual-polarized antenna aimed for wireless applications
operation in the frequency band of 5.15 GHz to 5.9 GHz. [7–12]. An exploded view of the dual-polarized microstrip
In each antenna element, two rectangular slots are used for antenna and a simplified equivalent circuit model for an
coupling the microstrip feed lines to the radiating patch. aperture coupled microstrip antenna are shown in Figure 1.
Ypatch
n1
Yap
Circular patch Slots in

ground
Ls plane 1
Rohacell Ls
Ver. feed
Substrate Wa Zin n2 Open stub
Hor. feed hf
(a) (b)
Circular patch Rohacell
hr
h
Spacer
Ground Dielectric
(c)
Figure 1: Configuration of the proposed dual-polarized aperture coupled circular patch antenna; (a) 3D view, (b) simplified equivalent
circuit model of an aperture coupled microstrip antenna, and (c) 2D view Rohacell: εr = 1.06, hr = 12 mm; substrate: εr f = 4.5, h f =
0.762 mm, h = 5.9 mm; vertical and horizontal apertures’ dimensions or feed slot (La × Wa ): 14 × 2 mm.
Rohacell
(radome)
Horizontal feed Vertical feed

(port 2) (port 1)
Metal
plate
(a) (b)
λ0 /4
λ0 /4
50 Ω
70.7 Ω
ρN
RL
ρ2 ρN
50 Ω 70.7 Ω ρ1
50 Ω
Z0 ΓIN ρ0 Z1 Z2 ZN
λ/4
(c) (d)
Figure 2: (a) 3D view of 8 × 8 array antenna with its ground plane. Rohacell (bottom: circular patches): εr = 1.06, hr = 12 mm; substrate
(top: slots, bottom: feed lines): εr f = 4.5, h f = 0.762 mm. (b) Feed structure of array antenna. (c) Quarter-wave matching transformer.
(d) N-section λ/4 transformer.
−5 9
−10
−15
8.5
Gain (dB)
−20
S (dB)
8
−25
−30
7.5
−35
−40 7
5 5.2 5.4 5.6 5.8 6 5 5.2 5.4 5.6 5.8 6
S11 S12 Vertical
S21 S22 Horizontal
(a) (b)
−20 0
−5
−30
−10
−40
−15
S (dB)
S (dB)
−50 −20
−25
−60
−30
−70
−35
5 5.2 5.4 5.6 5.8 6
5 5.2 5.4 5.6 5.8 6
S12 and S21 (Measurement) Measurement (S11) Simulation (S22)
S12 and S21 (Simulation) Measurement (S22) Simulation (S11)
(c) (d)
Figure 3: (a) Return loss and isolation versus frequency of one element of dual-polarized antenna element. (b) Gain versus frequency of one
element of dual polarized antenna element. (c) Isolation. (d) Return loss versus frequency of 8 × 8 array antenna.
Table 1: Wideband dual-polarized patch antenna specification. the bandwidth, and a radome. The input impedance of the
antenna at the center of the slot is given by [13, 14]
Frequency range 5.15–5.9 GHz
Peak gain 26 dBi n22
Zin = 2 − jZ0m cot βm Ls , (1)
Horizontal beamwidth 6◦ n1 Ypatch + Yap
Vertical beamwidth 7◦
where Ypatch is the patch admittance and Yap is the aperture
Front/back ratio Better than 28 dB admittance (Figure 1(b)). Z0m , βm , and Ls are the microstrip
Vertical: −11 dB (center frequency) line parameters in this equation. Also the coupling of the
SLL
Horizontal: −14 dB (center frequency) patch to the microstrip line is described by a transformer
Polarization (Dual) vertical or horizontal [14]. The dimensions of the element antenna such as slots,
VSWR 1.9 : 1 (max) feed lines, circular patch, and spaces between them are
Impedance 50 Ohms optimized with the use of IE3D to achieve best radiation
Mechanical Specification Length = width: 44 cm; depth: 4 cm characteristics, wide impedance bandwidth, and high iso-
lation between two ports. The optimized element antenna
has a circular patch with 11.89 mm radius positioned at
the bottom side of Rohacell. Furthermore, two 50 ohms
The antenna consists of only one substrate (Rogers TMM 4 microstrip feed lines (W = 1.5 mm, LV = 15 mm, and LH =
with dielectric constant εr = 4.5), an air layer for enhancing 23 mm) at the bottom side of the substrate (Rogers TMM 4
25 33
32
Front to back ratio (dB)

Front to back ratio (dB)
20
31
15
30
10
29
5 28
5 5.2 5.4 5.6 5.8 6 5 5.2 5.4 5.6 5.8 6
Vertical Vertical
Horizontal Horizontal
(a) (b)
Figure 4: (a) Simulated front-to-back ratio versus frequency of 8 × 8 array antenna without plate at the back of antenna. (b) Measured
front-to-back ratio versus frequency of 8 × 8 array antenna with metal plate at the back of antenna.
27 26
26 25
24
Gain (dB)
25
Gain (dB)
23
24
22
23
21
5 5.2 5.4 5.6 5.8 6
22
5 5.2 5.4 5.6 5.8 6 Frequency (GHz)
Frequency (GHz) Vertical
Horizontal
Vertical
Horizontal
(a) (b)
Figure 5: Gain versus frequency of 8 × 8 array antenna: (a) simulated and (b) measured.
with h f = 0.762 mm, εr f = 4.5) are electromagnetically 3. Array Antenna

coupled to circular patch through two rectangular slot
apertures in the common ground plane. As shown in To obtain the desired radiation pattern characteristics, an
Figure 2, in order to reduce the antenna back lobes, a metallic 8 × 8 planar microstrip slot array antenna is designed
plate is located at the back of antenna, for example, 22 mm (Figure 2(a)). The bottom side of substrate consists of the
from the bottom of the antenna structure. Additionally, feeding network which is designed to give equal amplitude
Figure 3(a) shows the simulated return loss for two ports and phase to each element (Figure 2(b)). Additionally,
(S11 and S22 ) versus frequency for one element antenna and by using T-junction design and a quarter-wave matching
Figure 3(b) shows simulated gain against frequency for one transformer (Figure 2(c)), the feeds are matched to 50 ohms
element. As depicted in Figure 3, in the desired bandwidth feed line [15, 16]. To provide a match, the transformer
(5.1–5.9) return loss for both polarizations is more than characteristic impedance Z1 should be Z1 = Rin Z0 , where
15 dB and the isolation between two ports (S12 ) is better than Z0 is the characteristic impedance of the input transmission
35 dB. line and Rin is the input impedance of the antenna. The
0 0
−10 −10
Normalized gain (dB)

−20 −20
−30 −30
−40 −40
−50 −50
−60 −60
−90 −50 0 50 90 −90 −50 0 50 90
θ θ
H plane H plane
E plane E plane
(a) (b)
Figure 6: Simulated antenna far-field radiation pattern at 5.5 GHz: (a) vertical and (b) horizontal.
0 0
−10 −10
−20 −20
−30 −30
−40 −40
−50 −50
−60 −60
−90−80 −60 −40 −20 0 20 40 60 80 90 −90 −50 0 50 90
θ θ
H plane H plane
E plane E plane
(a) (b)
Figure 7: Measured antenna far-field radiation pattern at 5.5 GHz: (a) vertical and (b) horizontal.
transformer is usually another transmission line with the antenna size, its gain, beamwidth, side lobe level, and front-
desired characteristic impedance (Figure 2(d)). The spaces to-back ratio are summarized in Table 1.
between elements are set at 50 mm for better radiation
characteristics. The simulated and measured return loss (S11 ) 4. Conclusions
and isolation (S21 ) of 8 × 8 dual-polarized microstrip patch
slot array antenna are illustrated in Figures 3(c) and 3(d). This paper has reported the design of a low-cost high-gain
Furthermore, the metal plate at the back of array antenna dual-polarized patch array antenna for WiMAX applications
reduces the front-to-back ratio about −20 dB, as can be in the 5.15–5.9 GHz frequency band. The antenna has an
seen in Figure 4. Likewise, the gain of the array antenna in approximately bandwidth of 14% and the peak gain of 26 dBi
different frequencies is demonstrated in Figure 5. Moreover, for both polarizations. The design has been achieved with
the simulated and measured E and H plane far-field radiation the use of commercial software packages AWR Microwave
patterns of the array antenna at center frequency are shown Office and Zeland IE3D. The design process aimed at best
in Figures 6 and 7. Finally, all vital parameters such as return losses and fine quality radiation characteristics over
the assumed frequency band. The designed antenna has an

impedance bandwidth of approximately 14% and the peak
gain of approximately 26 dBi for both polarizations. This
performance has been confirmed experimentally.
References
[1] C. A. Balanis, Antenna Theory, Analysis and Design, John Wiley
& Sons, 3rd edition, 2005.
[2] K. L. Wong, Planar Antennas for Wireless Communications,
John Wiley & Sons, 2003.
[3] S. B. Chen, Y. C. Jiao, W. Wang, and F. S. Zhang, “Modified T-
shaped planar monopole antennas for multiband operation,”
IEEE Transactions on Microwave Theory and Techniques, vol.
54, no. 8, pp. 3267–3270, 2006.
[4] W. C. Liu and C. F. Hsu, “Dual-band CPW-fed Y-shaped
monopole antenna for PCS/WLAN application,” Electronics
Letters, vol. 41, no. 7, pp. 390–391, 2005.
[5] W. C. Liu, “Broadband dual-frequency meandered CPW-fed
monopole antenna,” Electronics Letters, vol. 40, no. 21, pp.
1319–1320, 2004.
[6] J. Y. Li, J. L. Guo, Y. B. Gan, and Q. Z. Liu, “The tri-
band performance of sleeve dipole antenna,” Journal of
Electromagnetic Waves and Applications, vol. 19, no. 15, pp.
2081–2092, 2005.
[7] K. M. Z. Shams, M. Ali, and H. S. Hwang, “A planar induc-
tively coupled bow-tie slot antenna for WLAN application,”
Journal of Electromagnetic Waves and Applications, vol. 20, no.
7, pp. 861–871, 2006.
[8] C. Y. Wu, S. H. Yeh, and T. H. Lu, “Novel high gain
metamaterial antenna radome for WiMAX operation in the
5.8-GHz band,” in IEEE Antennas and Propagation Society
International Symposium (AP-S ’07), pp. 3488–3491, June
2007.
[9] L. N. Zhangl, S. S. Zhongl, and X. L. Liang, “Dual-band dual-
polarized hybrid antenna array,” in PIERS Proceedings, Xi’an,
China, March 2010.
[10] S. D. Targonski, R. B. Waterhouse, and D. M. Pozar, “Design of
wide-band aperture-stacked patch microstrip antennas,” IEEE
1245–1251, 1998.
[11] D. M. Pozar and S. D. Targonski, “A shared-aperture dual-
band dual-polarized microstrip array,” IEEE Transactions on
[12] H. J. Song and M. E. Bialkowski, “Ku-band 16 × 16 planar
array with aperture-coupled microstrip-patch elements,” IEEE
Antennas and Propagation Magazine, vol. 40, no. 5, pp. 25–29,
1998.
[13] G. Vetharatnam, C. B. Kuan, and C. H. Teik, “Combined
feed network for a shared-aperture dual-band dual-polarized
array,” IEEE Antennas and Wireless Propagation Letters, vol. 4,
no. 1, pp. 297–299, 2005.
[14] R. Garg and P. Bhartia, Microstrip Antenna Design Handbook,
Artech House, 2001.
[15] B. N. Das and K. K. Joshi, “Impedance of a radiating slot in
the ground plane of a microstripline,” IEEE Transactions on
[16] J. R. James and P. S. Hall, Handbook of Microstrip Antennas,
IEE Electromagnetic Waves, Peter Peregrinus, 1989.

Microstrip Antennas PDF

Загружено:

Сведения о документе

Оригинальное название

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Microstrip Antennas PDF

Загружено:

Авторское право:

Доступные форматы

International Journal of Antennas and Propagation

Series-Fed Microstrip Array Antenna with Circular Polarization, Tuan-Yung Han

Microstrip Patch Antenna Bandwidth Enhancement Using AMC/EBG Structures, R. C. Hadarig,

High-Performance Computational Electromagnetic Methods Applied to the Design of Patch Antenna

Jan Eichler, Pavel Hazdra, Miloslav Capek, and Milos Mazanek

Correspondence should be addressed to Pavel Hazdra, hazdrap@fel.cvut.cz

Received 9 August 2011; Revised 10 January 2012; Accepted 13 January 2012

Academic Editor: Charles Bunting

1. Introduction Equation (1) is usually treated within the method of

dielectrics. Our developed TCM tool [14] has the following

In [6] a novel theory able to rigorously calculate radiated

(b) Overlapping (Self) Elements. As known from the method 0.005

where T = T is a triangular area. Using simplex coordinates 3

20 an approximate analytical solution is available and in [25] we

25 where QJ is calculated from the currents using (11). Four

Figure 7: Relative error of the Q factor as a function of triangular

5. Applications: Fractal Antennas

Figure 14: Second higher mode J3 (currents and charges).

100 Using proper feeding techniques (like the L-probe) allows us

50 with parameter sweeps or even optimization, so we will be

Lingfei Mo and Chunfang Qin

Correspondence should be addressed to Lingfei Mo, lfmo@iipc.zju.edu.cn

Received 15 August 2011; Revised 9 December 2011; Accepted 28 December 2011

Academic Editor: Seong-Youp Suh

1. Introduction because the metallic surface boundary changes the radiation

CPW open stub 2

matching is dominated by the imaginary part. Under this

Chip reactance conjugate −10

Reactance matching point −20

Resistance matching point Antenna resistance

Power of the reader (dB)

Figure 9: Power bandwidth of the tag antenna.

4 Table 2: Tag performance testing results with the tag attached on

diﬀerent metal plate.

−20 300 60 −20 300 60

210 150 210 150

[17] Y. Um, U. Kim, and J. Choi, “Design of a compact CPW-fed

Correspondence should be addressed to Yong Liu, fatufo@bit.edu.cn

Received 12 August 2011; Revised 23 November 2011; Accepted 2 January 2012

Academic Editor: Zhongxiang Q. Shen

1. Introduction wave power increases and radiation power decreases with

[19]. It is also well known that antenna array is an eﬀective

Stripline feeding network

Microstrip feeding network

Figure 2: Radiation patterns of microstrip monopole only, mi-

Figure 3: Dual-polarized microstrip antenna array.

Frequency (GHz) 1.5

Figure 7: The structure of the DCWS.

is proposed by our group to meet this requirement [36].

Figure 9: The VSWR of sum and diﬀerence port of the monopulse

(b) The photograph of the monopulse circular-polarized

Figure 8: Ka-band monopulse circular-polarized dielectric com-

A Ka-band monopulse circular-polarized dielectric com-

Figure 13: The gain of the high integrate broad-band microstrip

Table 3: The measured data of the monopulse circular-polarized

1.8 4.1. High Integrate Broad-Band Microstrip Antenna Array

circuit board (MPCB) technology, the microstrip antennas

(a) E-plane (b) H-plane

integration with circuit, it is diﬃcult to be processed in

Workshop on Metamaterials, pp. 351–354, Nanjing, China,

Guido Valerio,1 Simona Mazzocchi,2 Alessandro Galli,2

Correspondence should be addressed to Guido Valerio, guido.valerio@univ-rennes1.fr

Received 15 August 2011; Revised 26 November 2011; Accepted 6 December 2011